MiMo V2.6 Pro pricing: real cost per million tokens
What Xiaomi's MiMo V2.6 Pro really costs across 11 plans from 3 providers: monthly fees, usable allowances and real per-token prices, against the official API list price and independent leaderboards.
Data updated · How real prices are computed
As of 2026-10-08, the cheapest way to use MiMo V2.6 Pro is Command Code GOAT at a real cost of about $0.00936 per million tokens ($10/month for roughly 1.069B tokens a month). We track 11 plans and API channels that serve MiMo V2.6 Pro. Official API list price per million tokens: $0.0036 cached input, $0.435 input, $0.87 output — about $0.0187 per million tokens on a standard workload mix.
- Lowest real price
- $0.00936/MTokCommand Code GOAT
- API list price
- $0.0187/MTokstandard workload mix
- Most monthly tokens
- 7.93BMiMo Token Plan Max (Night 0.8x)
- Plans & APIs
- 1110 subscriptions · 1 metered API
Plot MiMo V2.6 Pro on the price–capability chart →
MiMo V2.6 Pro plans and real prices
Real price = monthly fee ÷ tokens you can actually use in a month. “vs API” shows how many times cheaper than the official API list price (standard workload mix) the plan is. “Evidence” opens the full calculation and sources for that row.
| Plan | Provider | Billing | Monthly fee | Monthly tokens | Real $/MTok | vs API | Confidence | Data date | |
|---|---|---|---|---|---|---|---|---|---|
| Command Code GOAT | Command Code | Subscription | $10/mo | 1.069B | $0.00936 | 2× | Medium | Evidence ↓ | |
| OpenCode Go | OpenCode | Subscription | $10/mo | 0.8B | $0.01249 | 1.5× | Medium | Evidence ↓ | |
| MiMo Token Plan Max (Night 0.8x) | Xiaomi | Subscription | $100 (¥659)/mo | 7.93B | $0.01261 | 1.5× | Medium | Evidence ↓ | |
| MiMo Token Plan Pro (Night 0.8x) | Xiaomi | Subscription | $50 (¥329)/mo | 3.675B | $0.01361 | 1.4× | Medium | Evidence ↓ | |
| MiMo Token Plan Standard (Night 0.8x) | Xiaomi | Subscription | $16 (¥99)/mo | 1.064B | $0.01504 | 1.2× | Medium | Evidence ↓ | |
| MiMo Token Plan Lite (Night 0.8x) | Xiaomi | Subscription | $6 (¥39)/mo | 0.397B | $0.01511 | 1.2× | Medium | Evidence ↓ | |
| MiMo Token Plan Max (Day) | Xiaomi | Subscription | $100 (¥659)/mo | 6.344B | $0.01576 | 1.2× | Medium | Evidence ↓ | |
| MiMo Token Plan Pro (Day) | Xiaomi | Subscription | $50 (¥329)/mo | 2.94B | $0.01701 | 1.1× | Medium | Evidence ↓ | |
| MiMo V2.6 Pro API | Xiaomi | Metered API | — | — | $0.01872 | — | High | Evidence ↓ | |
| MiMo Token Plan Standard (Day) | Xiaomi | Subscription | $16 (¥99)/mo | 0.851B | $0.0188 | 1× | Medium | Evidence ↓ | |
| MiMo Token Plan Lite (Day) | Xiaomi | Subscription | $6 (¥39)/mo | 0.317B | $0.01893 | 1× | Medium | Evidence ↓ |
Behind each number
For every plan: the data date, the exact calculation, the quota basis, how leaderboard scores were mapped, and the original adoption evidence with its sources.
Command Code GOATCommand Code$0.00935891/MTok
- Data date
- 2026-09-30 · official source date
- Real price / MTok
- $0.00935891
- Monthly tokens
- 1,068,500,000 (1.069B)
- Quota confidence
- Medium
- Calculation
- $10 / month ÷ 1,068,500,000 tokens × 1,000,000 ≈ $0.00935891 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://commandcode.ai/docs/plans/goat official per-model allowance and three-part pricing; https://commandcode.ai/pricing; $10->$70 credits; code-subscriptions-round1-2026-09-06.json; goat-opencode-catalogs-round3-2026-09-30.json
Added 1.0685B: min(shared monthly pool $70, model allowance $20) ÷ project standard-workload weighted price; official three-part pricing; New models default $20. Processing fee ignored; allowances of different models in the same plan are not additive; no usage-panel token+% screenshot — computed from official absolute credits and the price list
https://commandcode.ai/docs/plans/goathttps://commandcode.ai/pricing;code-subscriptions-round1-2026-09-06.jsongoat-opencode-catalogs-round3-2026-09-30.jsonBrowse the public evidence archive
OpenCode GoOpenCode$0.0124938/MTok
- Plan
- OpenCode Go · OpenCode
- Data date
- 2026-09-30 · official source date
- Real price / MTok
- $0.0124938
- Monthly tokens
- 800,400,000 (0.8B)
- Quota confidence
- Medium
- Calculation
- $10 / month ÷ 800,400,000 tokens × 1,000,000 ≈ $0.0124938 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://opencode.ai/docs/go/ official per-model Usage and three-part pricing; opencode-go-round5-2026-09-06.json; goat-opencode-catalogs-round3-2026-09-30.json
Added 800.4M: min(shared monthly pool $60, model Usage $15) ÷ project standard-workload weighted price; single official tier; Usage $15; same Usage tier as v2.5-pro. The official request counts serve only as a cross-check and are no longer the primary allowance figure; allowances of different models in the same plan are not additive
https://opencode.ai/docs/go/goat-opencode-catalogs-round3-2026-09-30.jsonopencode-go-round5-2026-09-06.jsonBrowse the public evidence archive
MiMo Token Plan Max (Night 0.8x)Xiaomi$0.0126103/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0126103
- Monthly tokens
- 7,930,000,000 (7.93B)
- Quota confidence
- Medium
- Calculation
- $100 / month (¥659) ÷ 7,930,000,000 tokens × 1,000,000 ≈ $0.0126103 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 7.93B: monthly pool 82B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); night 00:00-08:00 (Beijing) consumption ×0.8 — the same credits buy 25% more tokens; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 3.22B — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Max (Night 0.8x) $100 list price (previously domestic ¥659÷6.7787≈$97.22); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo Token Plan Pro (Night 0.8x)Xiaomi$0.0136054/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0136054
- Monthly tokens
- 3,675,000,000 (3.675B)
- Quota confidence
- Medium
- Calculation
- $50 / month (¥329) ÷ 3,675,000,000 tokens × 1,000,000 ≈ $0.0136054 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 3.675B: monthly pool 38B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); night 00:00-08:00 (Beijing) consumption ×0.8 — the same credits buy 25% more tokens; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 1.492B — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Pro (Night 0.8x) $50 list price (previously domestic ¥329÷6.7787≈$48.53); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo Token Plan Standard (Night 0.8x)Xiaomi$0.0150376/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0150376
- Monthly tokens
- 1,064,000,000 (1.064B)
- Quota confidence
- Medium
- Calculation
- $16 / month (¥99) ÷ 1,064,000,000 tokens × 1,000,000 ≈ $0.0150376 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 1.064B: monthly pool 11B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); night 00:00-08:00 (Beijing) consumption ×0.8 — the same credits buy 25% more tokens; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 432M — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Standard (Night 0.8x) $16 list price (previously domestic ¥99÷6.7787≈$14.60); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo Token Plan Lite (Night 0.8x)Xiaomi$0.0151134/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0151134
- Monthly tokens
- 397,000,000 (0.397B)
- Quota confidence
- Medium
- Calculation
- $6 / month (¥39) ÷ 397,000,000 tokens × 1,000,000 ≈ $0.0151134 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 397M: monthly pool 4.1B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); night 00:00-08:00 (Beijing) consumption ×0.8 — the same credits buy 25% more tokens; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 161M — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Lite (Night 0.8x) $6 list price (previously domestic ¥39÷6.7787≈$5.75); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo Token Plan Max (Day)Xiaomi$0.0157629/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0157629
- Monthly tokens
- 6,344,000,000 (6.344B)
- Quota confidence
- Medium
- Calculation
- $100 / month (¥659) ÷ 6,344,000,000 tokens × 1,000,000 ≈ $0.0157629 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 6.344B: monthly pool 82B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); daytime baseline consumption tier; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 2.576B — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Max (Day) $100 list price (previously domestic ¥659÷6.7787≈$97.22); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo Token Plan Pro (Day)Xiaomi$0.0170068/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0170068
- Monthly tokens
- 2,940,000,000 (2.94B)
- Quota confidence
- Medium
- Calculation
- $50 / month (¥329) ÷ 2,940,000,000 tokens × 1,000,000 ≈ $0.0170068 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 2.94B: monthly pool 38B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); daytime baseline consumption tier; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 1.194B — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Pro (Day) $50 list price (previously domestic ¥329÷6.7787≈$48.53); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo V2.6 Pro APIXiaomi$0.018717/MTok
- Plan
- MiMo V2.6 Pro API · Xiaomi
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.018717
- Quota confidence
- High
- Calculation
- Metered API · public token prices weighted by the project standard workload.
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
Identical price across the OpenRouter/GOAT/OpenCode channels; Xiaomi's official pricing page does not list a V2.6 tier; mimo-v26-grok47-catalogs-round1-2026-09-22.json
List price cached 0.0036/in 0.435/out 0.87 × project standard workload 97.0%/2.50%/0.50%
mimo-v26-grok47-catalogs-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo Token Plan Standard (Day)Xiaomi$0.0188014/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0188014
- Monthly tokens
- 851,000,000 (0.851B)
- Quota confidence
- Medium
- Calculation
- $16 / month (¥99) ÷ 851,000,000 tokens × 1,000,000 ≈ $0.0188014 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 851M: monthly pool 11B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); daytime baseline consumption tier; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 346M — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Standard (Day) $16 list price (previously domestic ¥99÷6.7787≈$14.60); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo Token Plan Lite (Day)Xiaomi$0.0189274/MTok
- Data date
- 2026-09-22 · official source date
- Real price / MTok
- $0.0189274
- Monthly tokens
- 317,000,000 (0.317B)
- Quota confidence
- Medium
- Calculation
- $6 / month (¥39) ÷ 317,000,000 tokens × 1,000,000 ≈ $0.0189274 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.0036 / $0.435 / $0.87 · standard-load blended ≈ $0.0187/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (1 configuration)
Terminal-Bench 4.0 (AA): Low (1 configuration)
Code Arena | WebDev Overall: Low (1 configuration)
DeepSWE v1.1: Low (1 configuration)Mapping confidence: how directly each score refers to the model served here.
Adoption evidence · source and rationale (translated from the Chinese record)
https://mimo.mi.com/docs Token Plan official tiers/Credits pool/burn rate/night 0.8×/88% of list price for the first purchase; mimo-token-plan-round1-2026-09-22.json
Added 317M: monthly pool 4.1B Credits ÷ unified standard-workload blended burn 12.925 credits/token (this model cached/input/output=2.5/300/600 credits/token); daytime baseline consumption tier; the plan covers 4 text models v2.6-pro/v2.6-flash/v2.5-pro/v2.5 (2026-09-22 doc update + maintainer-panel corroboration); stops at depletion, no overdraft; allowances of different models in the same plan are not additive (assuming the shared Credits pool is used up on a single model); panel corroboration: on 2026-09-22 the official burn rate implied a single-day deduction of 2.347B vs the panel's actual 2.191B (−6.6%, explained by night 0.8× and timezone attribution) — panel-level corroboration of the burn rate; comparison: under the OpenCode harness single-day measured mix (cache read 92.04% / input 6.08% / output 1.88%) it would be 129M — not adopted: the ruled on 2026-09-23 that the low cache was caused by the OpenCode harness, and a two-question 35.72M tok measurement on another client hit cache 95.0%, so standard workload applies; evidence mimo-token-plan-panel-round2-2026-09-23.json, mimo-client-sample-round3-2026-09-23.json; same-name domestic and overseas tiers merged into one point: monthly fee and unit price use the international MiMo Token Plan Lite (Day) $6 list price (previously domestic ¥39÷6.7787≈$5.75); price/currency still record the actual domestic price paid; the allowance stays on the domestic-tier basis — the overseas same-name tier's absolute tokens are unmeasured; the merge is for price-tier display only
https://mimo.mi.com/docsmimo-client-sample-round3-2026-09-23.jsonmimo-token-plan-panel-round2-2026-09-23.jsonmimo-token-plan-round1-2026-09-22.jsonBrowse the public evidence archive
MiMo V2.6 Pro benchmark scores
Highest archived configuration per leaderboard; ranks are among all models each board archives in this dataset. Scores are never mixed across boards.
| Leaderboard | Metric | Score | Rank | Configuration | Effort | Harness | Configs | Snapshot |
|---|---|---|---|---|---|---|---|---|
| Artificial Analysis Intelligence Index v4.3 | Intelligence Index | 46.32 | #12 / 60 | MiMo-V2.6-Pro | — | — | 1 | 2026-10-01 |
| Terminal-Bench 4.0 (AA) | Resolution Rate % | 34.85 | #13 / 53 | Artificial Analysis - MiMo-V2.6-Pro | — | Artificial Analysis | 1 | 2026-10-01 |
| Code Arena | WebDev Overall | Arena Score | 1,619 | #18 / 54 | mimo-v2.6-pro | — | — | 1 | 2026-09-30 |
| DeepSWE v1.1 | Pass@1 % | 71.9 self-reported | #6 / 33 | mini-swe-agent - mimo-v2-6-pro | — | mini-swe-agent | 1 | 2026-09-03 |
Every archived configuration
Artificial Analysis Intelligence Index v4.3 1 configuration
Terminal-Bench 4.0 (AA) 1 configuration
| Configuration | Score | Score interval | Harness / effort / mode | Source task cost (mean / median) | Checked | Links |
|---|---|---|---|---|---|---|
| Artificial Analysis - MiMo-V2.6-ProAA's own-harness independent measurement of Terminal-Bench 4.0 (the terminalBench40 field in the models-board payload, fraction×100); scored separately from the official tbench.ai board. | 34.8485 | — | Artificial Analysis / — / — | — / — | 2026-10-01 | Source · Archive |
Code Arena | WebDev Overall 1 configuration
DeepSWE v1.1 1 configuration
| Configuration | Score | Score interval | Harness / effort / mode | Source task cost (mean / median) | Checked | Links |
|---|---|---|---|---|---|---|
| mini-swe-agent - mimo-v2-6-pro [vendor self-report]mini-swe-agent - mimo-v2-6-proThe official model card (MIT), section 3 Evaluation Results table, DeepSWE v1.1 row: Pro 71.9 / Flash 67.9 / V2.5 Pro 19.0. The same model has three mutually contradictory DeepSWE figures; per the record-all-choose-none rule all three are registered: ①the model card 71.9 (adopted for this tier — the same model-card basis as DeepSeek V4.1 Flash's self-report); ②the official release page mimo.mi.com/docs/en-US/news/latest/v2-6 records an RL curve 58.4→72.6; ③the third-party orcarouter read the Xiaomi Live panel endpoint 72.57 (panel captioned mini-swe-agent, avg@3). The differences come from checkpoint selection, rounding and avg@k; ② and ③ were not adopted. The official board has no MiMo row. | 71.9 vendor self-report | — | mini-swe-agent / — / — | — / — | 2026-09-22 | Source · Archive |
These are reference mappings, not benchmarks of a subscription or API channel. Harness and quota-measurement effort alignment are unverified. Source task costs are not subscription task costs.
- Exact served-model reference; quota-measurement effort is unverified.
- Exact served-model reference only; product harness and quota-measurement effort are unverified. Not a benchmark measurement of this subscription or API channel.
- Exact served-model reference only; product harness and quota-measurement effort are unverified. Not a benchmark measurement of this subscription or API channel. Vendor self-reported score, not an official leaderboard run.
MiMo V2.6 Pro official API pricing
| Cached input | Input | Output |
|---|---|---|
| $0.0036 | $0.435 | $0.87 |
Frequently asked questions
How much does MiMo V2.6 Pro really cost per million tokens?
Measured as monthly fee ÷ tokens you can actually use, MiMo V2.6 Pro ranges from $0.00936 (Command Code GOAT) to $0.01893 (MiMo Token Plan Lite (Day)) per million tokens.
Which subscription gives the most MiMo V2.6 Pro usage?
MiMo Token Plan Max (Night 0.8x) allows roughly 7.93B tokens of MiMo V2.6 Pro per month for $100 (¥659).
What is the official API price of MiMo V2.6 Pro?
Official API list price per million tokens: $0.0036 cached input, $0.435 input, $0.87 output — about $0.0187 per million tokens on a standard workload mix.
How does MiMo V2.6 Pro score on independent leaderboards?
MiMo V2.6 Pro has scores on 4 leaderboards; its strongest placement is 71.9 (Pass@1 %) on DeepSWE v1.1, #6 of 33 models tracked.
Who makes MiMo V2.6 Pro, and where can I use it?
MiMo V2.6 Pro (model ID mimo-v2.6-pro) is made by Xiaomi and is available through Command Code, OpenCode, Xiaomi.
More Xiaomi models
| Model | Lowest real $/MTok | Cheapest plan | Plans |
|---|---|---|---|
| MiMo V2.5 | $0.00127 | OpenCode Go | 10 |
| MiMo V2.6 Flash | $0.00127 | OpenCode Go | 11 |
| MiMo V2.5 Pro | $0.00936 | Command Code GOAT | 10 |
| MiMo V2.6 Pro UltraSpeed | $0.1872 | MiMo V2.6 Pro UltraSpeed API | 2 |