Step 3.7 Flash pricing: real cost per million tokens
What StepFun's Step 3.7 Flash really costs across 5 plans from 2 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 Step 3.7 Flash is Step Plan Max (¥699) at a real cost of about $0.0012 per million tokens ($103.1/month for roughly 85.883B tokens a month). We track 5 plans and API channels that serve Step 3.7 Flash. Official API list price per million tokens: ¥0.27 cached input, ¥1.35 input, ¥8.1 output — about $0.0496 per million tokens on a standard workload mix.
- Lowest real price
- $0.0012/MTokStep Plan Max (¥699)
- API list price
- $0.0496/MTokstandard workload mix
- Most monthly tokens
- 85.883BStep Plan Max (¥699)
- Plans & APIs
- 55 subscriptions · 0 metered APIs
Plot Step 3.7 Flash on the price–capability chart →
Step 3.7 Flash 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 | |
|---|---|---|---|---|---|---|---|---|---|
| Step Plan Max (¥699) | StepFun | Subscription | $103.1/mo | 85.883B | $0.0012 | 41.3× | Medium | Evidence ↓ | |
| Step Plan Pro (¥199) | StepFun | Subscription | $29.36/mo | 17.177B | $0.00171 | 29× | Medium | Evidence ↓ | |
| Step Plan Plus (¥99) | StepFun | Subscription | $14.6/mo | 3.435B | $0.00425 | 11.7× | Medium | Evidence ↓ | |
| Step Plan Mini (¥49) | StepFun | Subscription | $7.23/mo | 0.859B | $0.00842 | 5.9× | Medium | Evidence ↓ | |
| Command Code GOAT | Command Code | Subscription | $10/mo | 0.404B | $0.02478 | 2× | 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.
Step Plan Max (¥699)StepFun$0.00120067/MTok
- Plan
- Step Plan Max (¥699) · StepFun
- Data date
- 2026-09-10 · official source date
- Real price / MTok
- $0.00120067
- Monthly tokens
- 85,883,000,000 (85.883B)
- Quota confidence
- Medium
- Calculation
- $103.1 / month (¥699) ÷ 85,883,000,000 tokens × 1,000,000 ≈ $0.00120067 / MTok
- Quota basis
Low-cache workload: 85% cache reads / 14.5% input / 0.5% output
Official API list (cached / in / out, per MTok): ¥0.27 / ¥1.35 / ¥8.1 · standard-load blended ≈ $0.0496/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (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://platform.stepfun.com/docs/zh/step-plan/overview official Credit monthly pool, 1M Credit=¥1; https://platform.stepfun.com/docs/zh/guides/pricing/details RMB three-part pricing; stepfun-step-plan-round1-2026-09-10.json; stepfun-step-plan-round3-2026-09-10.json; stepfun-step-plan-round4-2026-09-10.json; stepfun-step5-panel-round5-2026-09-21.json
118.994B→85.883B: Step's measured load cannot reach the standard 97% cache (2026-09-22 ruling); switched to the unified low-cache workload (conventions.lowCacheTokenMix 85%/14.5%/0.5%) blended price¥0.466/M; CN-site monthly 40000M Credit ÷ that price, 1M Credit=¥1, cached/input/output=¥0.27/1.35/8.1. The English $1≈7M basis differs from the CNY basis by 0% for 3.5 and is 3.2% low for 3.7 due to USD-price rounding; the Chinese exact basis is adopted.The old Coding Plan Prompt/5h table was not used; Studio's 40% creation allowance not added; step-3.5-flash-2603 shares 3.5's price and is not listed separately; step-router-v1 gets no standalone point. No usage-panel token+% or saturated measurement — computed from official absolute Credit and price list + the unified low-cache basis; exchange rate 1 USD = 6.7787 CNY (2026-09-04 CNY central parity rate; old rate 7.2); CNY monthly fee divided by the rate to get USD; https://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.html
https://platform.stepfun.com/docs/zh/step-plan/overviewhttps://platform.stepfun.com/docs/zh/guides/pricing/detailshttps://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.htmlstepfun-step-plan-round1-2026-09-10.jsonstepfun-step-plan-round3-2026-09-10.jsonstepfun-step-plan-round4-2026-09-10.jsonstepfun-step5-panel-round5-2026-09-21.jsonBrowse the public evidence archive
Step Plan Pro (¥199)StepFun$0.00170911/MTok
- Plan
- Step Plan Pro (¥199) · StepFun
- Data date
- 2026-09-10 · official source date
- Real price / MTok
- $0.00170911
- Monthly tokens
- 17,176,600,000 (17.177B)
- Quota confidence
- Medium
- Calculation
- $29.36 / month (¥199) ÷ 17,176,600,000 tokens × 1,000,000 ≈ $0.00170911 / MTok
- Quota basis
Low-cache workload: 85% cache reads / 14.5% input / 0.5% output
Official API list (cached / in / out, per MTok): ¥0.27 / ¥1.35 / ¥8.1 · standard-load blended ≈ $0.0496/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (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://platform.stepfun.com/docs/zh/step-plan/overview official Credit monthly pool, 1M Credit=¥1; https://platform.stepfun.com/docs/zh/guides/pricing/details RMB three-part pricing; stepfun-step-plan-round1-2026-09-10.json; stepfun-step-plan-round3-2026-09-10.json; stepfun-step-plan-round4-2026-09-10.json; stepfun-step5-panel-round5-2026-09-21.json
23.7989B→17.1766B: Step's measured load cannot reach the standard 97% cache (2026-09-22 ruling); switched to the unified low-cache workload (conventions.lowCacheTokenMix 85%/14.5%/0.5%) blended price¥0.466/M; CN-site monthly 8000M Credit ÷ that price, 1M Credit=¥1, cached/input/output=¥0.27/1.35/8.1. The English $1≈7M basis differs from the CNY basis by 0% for 3.5 and is 3.2% low for 3.7 due to USD-price rounding; the Chinese exact basis is adopted.The old Coding Plan Prompt/5h table was not used; Studio's 40% creation allowance not added; step-3.5-flash-2603 shares 3.5's price and is not listed separately; step-router-v1 gets no standalone point. No usage-panel token+% or saturated measurement — computed from official absolute Credit and price list + the unified low-cache basis; exchange rate 1 USD = 6.7787 CNY (2026-09-04 CNY central parity rate; old rate 7.2); CNY monthly fee divided by the rate to get USD; https://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.html
https://platform.stepfun.com/docs/zh/step-plan/overviewhttps://platform.stepfun.com/docs/zh/guides/pricing/detailshttps://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.htmlstepfun-step-plan-round1-2026-09-10.jsonstepfun-step-plan-round3-2026-09-10.jsonstepfun-step-plan-round4-2026-09-10.jsonstepfun-step5-panel-round5-2026-09-21.jsonBrowse the public evidence archive
Step Plan Plus (¥99)StepFun$0.00425132/MTok
- Plan
- Step Plan Plus (¥99) · StepFun
- Data date
- 2026-09-10 · official source date
- Real price / MTok
- $0.00425132
- Monthly tokens
- 3,435,300,000 (3.435B)
- Quota confidence
- Medium
- Calculation
- $14.6 / month (¥99) ÷ 3,435,300,000 tokens × 1,000,000 ≈ $0.00425132 / MTok
- Quota basis
Low-cache workload: 85% cache reads / 14.5% input / 0.5% output
Official API list (cached / in / out, per MTok): ¥0.27 / ¥1.35 / ¥8.1 · standard-load blended ≈ $0.0496/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (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://platform.stepfun.com/docs/zh/step-plan/overview official Credit monthly pool, 1M Credit=¥1; https://platform.stepfun.com/docs/zh/guides/pricing/details RMB three-part pricing; stepfun-step-plan-round1-2026-09-10.json; stepfun-step-plan-round3-2026-09-10.json; stepfun-step-plan-round4-2026-09-10.json; stepfun-step5-panel-round5-2026-09-21.json
4.7598B→3.4353B: Step's measured load cannot reach the standard 97% cache (2026-09-22 ruling); switched to the unified low-cache workload (conventions.lowCacheTokenMix 85%/14.5%/0.5%) blended price¥0.466/M; CN-site monthly 1600M Credit ÷ that price, 1M Credit=¥1, cached/input/output=¥0.27/1.35/8.1. The English $1≈7M basis differs from the CNY basis by 0% for 3.5 and is 3.2% low for 3.7 due to USD-price rounding; the Chinese exact basis is adopted.The old Coding Plan Prompt/5h table was not used; Studio's 40% creation allowance not added; step-3.5-flash-2603 shares 3.5's price and is not listed separately; step-router-v1 gets no standalone point. No usage-panel token+% or saturated measurement — computed from official absolute Credit and price list + the unified low-cache basis; exchange rate 1 USD = 6.7787 CNY (2026-09-04 CNY central parity rate; old rate 7.2); CNY monthly fee divided by the rate to get USD; https://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.html
https://platform.stepfun.com/docs/zh/step-plan/overviewhttps://platform.stepfun.com/docs/zh/guides/pricing/detailshttps://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.htmlstepfun-step-plan-round1-2026-09-10.jsonstepfun-step-plan-round3-2026-09-10.jsonstepfun-step-plan-round4-2026-09-10.jsonstepfun-step5-panel-round5-2026-09-21.jsonBrowse the public evidence archive
Step Plan Mini (¥49)StepFun$0.00841701/MTok
- Plan
- Step Plan Mini (¥49) · StepFun
- Data date
- 2026-09-10 · official source date
- Real price / MTok
- $0.00841701
- Monthly tokens
- 858,800,000 (0.859B)
- Quota confidence
- Medium
- Calculation
- $7.23 / month (¥49) ÷ 858,800,000 tokens × 1,000,000 ≈ $0.00841701 / MTok
- Quota basis
Low-cache workload: 85% cache reads / 14.5% input / 0.5% output
Official API list (cached / in / out, per MTok): ¥0.27 / ¥1.35 / ¥8.1 · standard-load blended ≈ $0.0496/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (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://platform.stepfun.com/docs/zh/step-plan/overview official Credit monthly pool, 1M Credit=¥1; https://platform.stepfun.com/docs/zh/guides/pricing/details RMB three-part pricing; stepfun-step-plan-round1-2026-09-10.json; stepfun-step-plan-round3-2026-09-10.json; stepfun-step-plan-round4-2026-09-10.json; stepfun-step5-panel-round5-2026-09-21.json
1.1899B→858.8M: Step's measured load cannot reach the standard 97% cache (2026-09-22 ruling); switched to the unified low-cache workload (conventions.lowCacheTokenMix 85%/14.5%/0.5%) blended price¥0.466/M; CN-site monthly 400M Credit ÷ that price, 1M Credit=¥1, cached/input/output=¥0.27/1.35/8.1. The English $1≈7M basis differs from the CNY basis by 0% for 3.5 and is 3.2% low for 3.7 due to USD-price rounding; the Chinese exact basis is adopted.The old Coding Plan Prompt/5h table was not used; Studio's 40% creation allowance not added; step-3.5-flash-2603 shares 3.5's price and is not listed separately; step-router-v1 gets no standalone point. No usage-panel token+% or saturated measurement — computed from official absolute Credit and price list + the unified low-cache basis; exchange rate 1 USD = 6.7787 CNY (2026-09-04 CNY central parity rate; old rate 7.2); CNY monthly fee divided by the rate to get USD; https://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.html
https://platform.stepfun.com/docs/zh/step-plan/overviewhttps://platform.stepfun.com/docs/zh/guides/pricing/detailshttps://www.xinhuanet.com/20260904/4c1ce1cbf39d416997cb9c237d6f5024/c.htmlstepfun-step-plan-round1-2026-09-10.jsonstepfun-step-plan-round3-2026-09-10.jsonstepfun-step-plan-round4-2026-09-10.jsonstepfun-step5-panel-round5-2026-09-21.jsonBrowse the public evidence archive
Command Code GOATCommand Code$0.024777/MTok
- Data date
- 2026-09-30 · official source date
- Real price / MTok
- $0.024777
- Monthly tokens
- 403,600,000 (0.404B)
- Quota confidence
- Medium
- Calculation
- $10 / month ÷ 403,600,000 tokens × 1,000,000 ≈ $0.024777 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): ¥0.27 / ¥1.35 / ¥8.1 · standard-load blended ≈ $0.0496/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (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 403.6M: 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
Step 3.7 Flash 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 | 19.48 estimate | #57 / 60 | Step 3.7 Flash | — | — | 1 | 2026-10-01 |
Every archived configuration
Artificial Analysis Intelligence Index v4.3 1 configuration
| Configuration | Score | Score interval | Harness / effort / mode | Source task cost (mean / median) | Checked | Links |
|---|---|---|---|---|---|---|
| Step 3.7 Flash [AA estimate]Step 3.7 FlashMarked intelligenceIndexIsEstimated=true in AA embedded payload; retained as estimate, not treated as fully measured. | 19.4813 AA estimate | — | — / — / — | — / — | 2026-10-01 | 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.
Step 3.7 Flash official API pricing
| Cached input | Input | Output |
|---|---|---|
| ¥0.27 | ¥1.35 | ¥8.1 |
Frequently asked questions
How much does Step 3.7 Flash really cost per million tokens?
Measured as monthly fee ÷ tokens you can actually use, Step 3.7 Flash ranges from $0.0012 (Step Plan Max (¥699)) to $0.02478 (Command Code GOAT) per million tokens.
Which subscription gives the most Step 3.7 Flash usage?
Step Plan Max (¥699) allows roughly 85.883B tokens of Step 3.7 Flash per month for $103.1.
What is the official API price of Step 3.7 Flash?
Official API list price per million tokens: ¥0.27 cached input, ¥1.35 input, ¥8.1 output — about $0.0496 per million tokens on a standard workload mix.
How does Step 3.7 Flash score on independent leaderboards?
Step 3.7 Flash has scores on 1 leaderboard; its strongest placement is 19.48 (Intelligence Index) on Artificial Analysis Intelligence Index v4.3, #57 of 60 models tracked.
Who makes Step 3.7 Flash, and where can I use it?
Step 3.7 Flash (model ID step-3.7-flash) is made by StepFun and is available through Command Code, StepFun.
More StepFun models
| Model | Lowest real $/MTok | Cheapest plan | Plans |
|---|---|---|---|
| Step 3.5 Flash | $0.0006 | Step Plan Max (¥699) | 5 |
| Step 5 Preview | $0.00364 | Step Plan Max (¥699) | 5 |