DeepSeek V4 Flash pricing: real cost per million tokens
What DeepSeek's DeepSeek V4 Flash really costs across 6 plans from 4 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 DeepSeek V4 Flash is Command Code GOAT at a real cost of about $0.00161 per million tokens ($10/month for roughly 6.211B tokens a month). We track 6 plans and API channels that serve DeepSeek V4 Flash. Official API list price per million tokens: $0.014 cached input, $0.44 input, $1.32 output — about $0.0312 per million tokens on a standard workload mix.
- Lowest real price
- $0.00161/MTokCommand Code GOAT
- API list price
- $0.0312/MTokstandard workload mix
- Most monthly tokens
- 19.243BOllama Max
- Plans & APIs
- 64 subscriptions · 2 metered APIs
Plot DeepSeek V4 Flash on the price–capability chart →
DeepSeek V4 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 | |
|---|---|---|---|---|---|---|---|---|---|
| Command Code GOAT | Command Code | Subscription | $10/mo | 6.211B | $0.00161 | 19.4× | Medium | Evidence ↓ | |
| OpenCode Go | OpenCode | Subscription | $10/mo | 3.106B | $0.00322 | 9.7× | Medium | Evidence ↓ | |
| Ollama Max | Ollama | Subscription | $100/mo | 19.243B | $0.0052 | 6× | Medium | Evidence ↓ | |
| Ollama Pro | Ollama | Subscription | $20/mo | 3.849B | $0.0052 | 6× | Medium | Evidence ↓ | |
| DeepSeek V4 Flash API Off-peak | DeepSeek | Metered API | — | — | $0.01559 | — | High | Evidence ↓ | |
| DeepSeek V4 Flash API Peak | DeepSeek | Metered API | — | — | $0.03118 | — | High | 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$0.00160999/MTok
- Data date
- 2026-09-30 · official source date
- Real price / MTok
- $0.00160999
- Monthly tokens
- 6,211,200,000 (6.211B)
- Quota confidence
- Medium
- Calculation
- $10 / month ÷ 6,211,200,000 tokens × 1,000,000 ≈ $0.00160999 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.014 / $0.44 / $1.32 · standard-load blended ≈ $0.0312/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (4 configurations)
Terminal-Bench 4.0 (AA): Low (3 configurations)
Code Arena | WebDev Overall: Low (2 configurations)
Artificial Analysis Coding Agent Index v1.5: 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; command-code-goat-deepseek-round2-2026-09-30.json
restored: official 'V4 Flash (latest)' allowance $60 (reverified 2026-09-30; removed 09-10 as confirmed offline); min(shared monthly pool $70, model allowance $60) / standard-workload blended price; off-peak cached/input/output=$0.003/$0.15/$0.60, peak 2x; official estimate 154,000 requests/mo. Processing fee ignored; per-model allowances within the plan do not add
https://commandcode.ai/docs/plans/goathttps://commandcode.ai/pricingcommand-code-goat-deepseek-round2-2026-09-30.jsonBrowse the public evidence archive
OpenCode GoOpenCode$0.00321999/MTok
- Plan
- OpenCode Go · OpenCode
- Data date
- 2026-09-30 · official source date
- Real price / MTok
- $0.00321999
- Monthly tokens
- 3,105,600,000 (3.106B)
- Quota confidence
- Medium
- Calculation
- $10 / month ÷ 3,105,600,000 tokens × 1,000,000 ≈ $0.00321999 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.014 / $0.44 / $1.32 · standard-load blended ≈ $0.0312/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (4 configurations)
Terminal-Bench 4.0 (AA): Low (3 configurations)
Code Arena | WebDev Overall: Low (2 configurations)
Artificial Analysis Coding Agent Index v1.5: 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-deepseek-round7-2026-09-30.json
restored: official Usage $30 (reverified 2026-09-30; removed 09-10 as confirmed offline); min(shared monthly pool $60, model Usage $30) / standard-workload blended price; off-peak cached/input/output=$0.003/$0.15/$0.60, peak 2x; official estimate 65,000 requests/mo. Per-model allowances within the plan do not add
https://opencode.ai/docs/go/opencode-go-deepseek-round7-2026-09-30.jsonBrowse the public evidence archive
Ollama MaxOllama$0.00519667/MTok
- Plan
- Ollama Max · Ollama
- Data date
- 2026-09-06 · official source date
- Real price / MTok
- $0.00519667
- Monthly tokens
- 19,243,100,000 (19.243B)
- Quota confidence
- Medium
- Calculation
- $100 / month ÷ 19,243,100,000 tokens × 1,000,000 ≈ $0.00519667 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.014 / $0.44 / $1.32 · standard-load blended ≈ $0.0312/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (4 configurations)
Terminal-Bench 4.0 (AA): Low (3 configurations)
Code Arena | WebDev Overall: Low (2 configurations)
Artificial Analysis Coding Agent Index v1.5: 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://ollama.com/pricing official usage credits and three-part pricing; https://ollama.com/blog/transparent-pricing; code-subscriptions-round1-2026-09-06.json; ollama-deepseek-v41-round1-2026-09-11.json
Added 19.2431B: shared monthly pool $300 ÷ project standard-workload weighted price; Off-Peak; Peak=2× (12:00–18:00 UTC Mon–Fri), same peak/valley basis as the project's OpenCode DeepSeek rows. Allowances of different models in the same plan are not additive (assuming the shared pool is used up on a single model); no usage-panel token+% screenshot — computed from official absolute credits and the price list
https://ollama.com/pricinghttps://ollama.com/blog/transparent-pricingcode-subscriptions-round1-2026-09-06.jsonollama-deepseek-v41-round1-2026-09-11.jsonBrowse the public evidence archive
Ollama ProOllama$0.00519669/MTok
- Plan
- Ollama Pro · Ollama
- Data date
- 2026-09-06 · official source date
- Real price / MTok
- $0.00519669
- Monthly tokens
- 3,848,600,000 (3.849B)
- Quota confidence
- Medium
- Calculation
- $20 / month ÷ 3,848,600,000 tokens × 1,000,000 ≈ $0.00519669 / MTok
- Quota basis
Standard workload: 97% cache reads / 2.5% input / 0.5% output
Official API list (cached / in / out, per MTok): $0.014 / $0.44 / $1.32 · standard-load blended ≈ $0.0312/MTok
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (4 configurations)
Terminal-Bench 4.0 (AA): Low (3 configurations)
Code Arena | WebDev Overall: Low (2 configurations)
Artificial Analysis Coding Agent Index v1.5: 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://ollama.com/pricing official usage credits and three-part pricing; https://ollama.com/blog/transparent-pricing; code-subscriptions-round1-2026-09-06.json; ollama-deepseek-v41-round1-2026-09-11.json
Added 3.8486B: shared monthly pool $60 ÷ project standard-workload weighted price; Off-Peak; Peak=2× (12:00–18:00 UTC Mon–Fri), same peak/valley basis as the project's OpenCode DeepSeek rows. Allowances of different models in the same plan are not additive (assuming the shared pool is used up on a single model); no usage-panel token+% screenshot — computed from official absolute credits and the price list
https://ollama.com/pricinghttps://ollama.com/blog/transparent-pricingcode-subscriptions-round1-2026-09-06.jsonollama-deepseek-v41-round1-2026-09-11.jsonBrowse the public evidence archive
DeepSeek V4 Flash API Off-peakDeepSeek$0.01559/MTok
- Data date
- 2026-09-05 · official source date
- Real price / MTok
- $0.01559
- Quota confidence
- High
- Calculation
- Metered API · public token prices weighted by the project standard workload.
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (4 configurations)
Terminal-Bench 4.0 (AA): Low (3 configurations)
Code Arena | WebDev Overall: Low (2 configurations)
Artificial Analysis Coding Agent Index v1.5: 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)
api-docs.deepseek.com
List price cached 0.007/in 0.22/out 0.66 × project standard workload 97.0%/2.50%/0.50%
DeepSeek V4 Flash API PeakDeepSeek$0.03118/MTok
- Data date
- 2026-09-05 · official source date
- Real price / MTok
- $0.03118
- Quota confidence
- High
- Calculation
- Metered API · public token prices weighted by the project standard workload.
- Benchmark mapping
- Artificial Analysis Intelligence Index v4.3: Medium (4 configurations)
Terminal-Bench 4.0 (AA): Low (3 configurations)
Code Arena | WebDev Overall: Low (2 configurations)
Artificial Analysis Coding Agent Index v1.5: 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)
api-docs.deepseek.com
List price cached 0.014/in 0.44/out 1.32 × project standard workload 97.0%/2.50%/0.50%
DeepSeek V4 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 | 34.33 | #34 / 60 | DeepSeek V4 Flash 0731 (max) | Max | — | 4 | 2026-10-01 |
| Terminal-Bench 4.0 (AA) | Resolution Rate % | 12.12 | #30 / 53 | Artificial Analysis - DeepSeek V4 Flash 0731 (max) | Max | Artificial Analysis | 3 | 2026-10-01 |
| Code Arena | WebDev Overall | Arena Score | 1,581 | #27 / 54 | deepseek-v4-flash-high | High | — | 2 | 2026-09-30 |
| Artificial Analysis Coding Agent Index v1.5 | Coding Agent Index | 38.74 | #19 / 19 | Codex - DeepSeek V4 Flash 0731 (max) | Max | Codex | 1 | 2026-10-01 |
| DeepSWE v1.1 | Pass@1 % | 53.32 | #24 / 33 | mini-swe-agent - deepseek-v4-flash (max) | Max | mini-swe-agent | 1 | 2026-09-03 |
Every archived configuration
Artificial Analysis Intelligence Index v4.3 4 configurations
| Configuration | Score | Score interval | Harness / effort / mode | Source task cost (mean / median) | Checked | Links |
|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 (max) | 34.3305 | — | — / Max / — | — / — | 2026-10-01 | Source · Archive |
| DeepSeek V4 Flash (high) | 24.384 | — | — / High / — | — / — | 2026-10-01 | Source · Archive |
| DeepSeek V4 Flash (max) | 24.1673 | — | — / Max / — | — / — | 2026-10-01 | Source · Archive |
| DeepSeek V4 Flash (non-reasoning) [AA estimate]DeepSeek V4 Flash (non-reasoning)Marked intelligenceIndexIsEstimated=true in AA embedded payload; retained as estimate, not treated as fully measured. | 18.8769 AA estimate | — | — / — / — | — / — | 2026-10-01 | Source · Archive |
Terminal-Bench 4.0 (AA) 3 configurations
| Configuration | Score | Score interval | Harness / effort / mode | Source task cost (mean / median) | Checked | Links |
|---|---|---|---|---|---|---|
| Artificial Analysis - DeepSeek V4 Flash 0731 (max)AA'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. | 12.1212 | — | Artificial Analysis / Max / — | — / — | 2026-10-01 | Source · Archive |
| Artificial Analysis - DeepSeek V4 Flash (high)AA'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. | 3.0303 | — | Artificial Analysis / High / — | — / — | 2026-10-01 | Source · Archive |
| Artificial Analysis - DeepSeek V4 Flash (max)AA'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. | 2.5253 | — | Artificial Analysis / Max / — | — / — | 2026-10-01 | Source · Archive |
Code Arena | WebDev Overall 2 configurations
| Configuration | Score | Score interval | Harness / effort / mode | Source task cost (mean / median) | Checked | Links |
|---|---|---|---|---|---|---|
| deepseek-v4-flash-high | 1,581 | 1,571.8 – 1,590.2 | — / High / — | — / — | 2026-10-01 | Source · Archive |
| deepseek-v4-flash-high-preview | 1,430 | 1,423.1 – 1,436.9 | — / High / — | — / — | 2026-10-01 | Source · Archive |
Artificial Analysis Coding Agent Index v1.5 1 configuration
| Configuration | Score | Score interval | Harness / effort / mode | Source task cost (mean / median) | Checked | Links |
|---|---|---|---|---|---|---|
| Codex - DeepSeek V4 Flash 0731 (max)Score belongs to this named agent/harness × model configuration; it is not a harness-independent model measurement. indexScore stored as percent (raw fraction × 100). | 38.7425 | — | Codex / Max / — | $0.0851 / — | 2026-10-01 | Source · Archive |
DeepSWE v1.1 1 configuration
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.
DeepSeek V4 Flash official API pricing
| Cached input | Input | Output |
|---|---|---|
| $0.014 | $0.44 | $1.32 |
Frequently asked questions
How much does DeepSeek V4 Flash really cost per million tokens?
Measured as monthly fee ÷ tokens you can actually use, DeepSeek V4 Flash ranges from $0.00161 (Command Code GOAT) to $0.03118 (DeepSeek V4 Flash API Peak) per million tokens.
Which subscription gives the most DeepSeek V4 Flash usage?
Ollama Max allows roughly 19.243B tokens of DeepSeek V4 Flash per month for $100.
What is the official API price of DeepSeek V4 Flash?
Official API list price per million tokens: $0.014 cached input, $0.44 input, $1.32 output — about $0.0312 per million tokens on a standard workload mix.
How does DeepSeek V4 Flash score on independent leaderboards?
DeepSeek V4 Flash has scores on 5 leaderboards; its strongest placement is 1,581 (Arena Score) on Code Arena | WebDev Overall, #27 of 54 models tracked.
Who makes DeepSeek V4 Flash, and where can I use it?
DeepSeek V4 Flash (model ID deepseek-v4-flash) is made by DeepSeek and is available through Command Code, DeepSeek, Ollama, OpenCode.
More DeepSeek models
| Model | Lowest real $/MTok | Cheapest plan | Plans |
|---|---|---|---|
| DeepSeek V4.1 Flash | $0.00161 | OpenCode Go | 7 |
| DeepSeek V4.1 Flash Fast | $0.00374 | Command Code GOAT | 1 |
| DeepSeek V4 Flash Vision Exp | $0.00483 | Command Code GOAT | 2 |
| DeepSeek V4 Pro | $0.01591 | Ollama Pro | 6 |
| DeepSeek V4 Flash Fast | $0.03885 | Command Code GOAT | 1 |