Models / DeepSeek-V4.1-Flash
DeepSeek-V4.1-Flash
DeepSeek-V4.1-Flash is best for writing, and agents and tool use; a strong pick at a modest price. The cheapest model of similar intelligence is MiMo-V2.6-Flash, at $0.18 per million tokens against $0.52.
Superseded. DeepSeek-V4.1-Flash-1017 is DeepSeek's newer model in this line, at $0.52 blended: see DeepSeek-V4.1-Flash-1017 or compare the two.
Results 16 Nov 2026 · 56 results · 3 sources
Test DeepSeek-V4.1-Flash on your prompt → It runs beside the models you could use instead, with every answer checked and costed.
- Price per million tokens
- $0.30 in · $1.20 out
DeepSeek's list price, 10 Nov 2026 · compare prices$0.01 cached input - Speed
- 369 tokens a second; on its test prompt the answer starts after 0.6 s, thinking time included
Artificial Analysis, on its usual API, at no reasoning; 231–369 across 2 settings - Intelligence Index
- 39.5
Artificial Analysis, at max reasoning; 24.7–39.5 across 2 settings - Spring Prompt overall
- 18th of 36 score 51 of 100 · how it works
- Developer
- DeepSeek, based in China
Weights published - Released
- 10 Sep 2026
1,048,576 tokens of context - Pricing and availability
- Half price off-peak ($0.15 / $0.60)
Where it stands among the models you could choose
Available models grouped by price, intelligence, speed and developer location, plus models with published weights. Retired models are excluded.
| Among | Intelligence Index | Spring Prompt overall | Price, blended | Speed |
|---|---|---|---|---|
| Models available today | 30th of 185Claude Opus 5.5 | 18th of 35Claude Fable 5.2 | $0.52 blended, 38th percentilecheapest: Mistral Nemo | 3rd of 65Nemotron 3.5 Super VL |
| Models priced $0.25–1 per million tokens | 4th of 61MiMo-V2.6-Pro | 1st of 5leads | sets this group | 2nd of 20Gemini 3.5 Flash-Lite |
| Models with similar intelligence | sets this group | 2nd of 6Gemini 3.8 Flash | $0.52 blended, 30th percentilecheapest: MiMo-V2.6-Flash | 1st of 12leads |
| Models writing 200 or more tokens a second | 4th of 14Muse Spark 1.3 | 2nd of 5Muse Spark 1.3 | $0.52 blended, 54th percentilecheapest: gpt-oss-20b | sets this group |
| Models from developers based in China | 9th of 75MiMo-V2.6-Pro | 3rd of 11Kimi K3.1 Pro | $0.52 blended, 47th percentilecheapest: Qwen3.5-9B | 1st of 23leads |
| Open-weights models | 8th of 104MiMo-V2.6-Pro | 2nd of 8Kimi K3.1 Pro | $0.52 blended, 56th percentilecheapest: Mistral Nemo | 2nd of 43Nemotron 3.5 Super VL |
The name under each rank is the leader of that group; green is the top quarter of the group and red the bottom quarter, by rank. Price is shown as a percentile: the share of the group that costs less, so lower is cheaper. "Sets this group" marks the measure the group is defined by. Spring Prompt overall is our score out of 100 across every source (how it works). Price is the developer's list price (or the typical OpenRouter host where we haven't read one) for three input tokens to one output; intelligence and speed are from Artificial Analysis (data sourced from Artificial Analysis); "similar intelligence" means within 4 points on its Intelligence Index. Developer locations are where each company is based, not where a model is served. Groups under 3 models are not ranked.
Measured cost per task
API cost recorded in our benchmark runs, excluding judging.
| Task | Benchmark | Cost per task |
|---|---|---|
| Product listing | CatalogBench | ≈ $0.006 |
Cheapest model that passes
On our benchmarks that define a pass for each task: what one passing answer cost in our run, failed attempts included, and whether a cheaper model passed.
- CatalogBench · cost per 1,000 accepted listings
DeepSeek-V4.1-Flash: $9.53 (107 of 168 attempts), 6 of 25. Cheapest: GPT-6 Luna at $0.70 (114 of 168 attempts). It passed at least as often.
Cost per attempt divided by the share that passed, at the run date's prices, excluding judging. Each model at its cheapest setting per passing answer. No estimate where a run had missing outputs, because their cost may not have been recorded.
How good is it, and for what?
Each use case ranks the available models its benchmarks measured, then those priced $0.25–1 per million tokens. The bar shows its percentile in that field, best to the right. Open a row for the results behind it.
Product listingsTurning a sparse product feed and photos into listings that can go live
| Benchmark | DeepSeek-V4.1-Flash | Rank among available | Best available |
|---|---|---|---|
| CatalogBenchCatalogBench: reliably publish-ready | 41.1% | 17th of 32 | GPT-6 Astra 75.0% |
| CatalogBenchCatalogBench (sales brief): reliably publish-ready | 23.2% | 11th of 32 | GPT-6 Astra 69.6% |
Decks from an analysisTurning a finished analysis into a deck you could present as it is
- DeckBench: has not measured this model.
User surveysPlanning a user survey and reading its results without being misled
- SurveyBench: has not measured this model.
Marketing planningPlanning a year of ad spend without overspending
- ROASBench: has not measured this model.
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
| Benchmark | DeepSeek-V4.1-Flash | Rank among available | Best available |
|---|---|---|---|
| Artificial AnalysisArtificial Analysis: Terminal-Bench 4.0 · reported | 26.8% | 22nd of 107 | Kimi K3.1 Pro 67.2% |
- tau2-bench: pending. Its latest results are dated 26 May 2026, before this model was released.
- Berkeley Function Calling Leaderboard (BFCL) V4: pending. Its latest results are dated 16 Dec 2025, before this model was released.
- OpenHands Index: pending. Its latest results are dated 30 Jun 2026, before this model was released.
- Microsoft STATE-Bench: pending. Its latest results are dated 29 May 2026, before this model was released.
- Vending-Bench 2: has not measured this model.
Professional workReal tasks from banking, consulting and law, business documents and freelance projects
- APEX-Agents: has not measured this model.
- GDP.pdf: has not measured this model.
- Remote Labor Index: has not measured this model.
Reasoning and knowledgeHard questions across science, maths and general knowledge
| Benchmark | DeepSeek-V4.1-Flash | Rank among available | Best available |
|---|---|---|---|
| Artificial AnalysisArtificial Analysis: Artificial Analysis Intelligence Index · reported | 39.5 | 30th of 185 | Claude Opus 5.5 57.6 |
| Artificial AnalysisArtificial Analysis: Humanity's Last Exam · reported | 39.2% | 47th of 186 | Claude Opus 5.5 61.4% |
WritingWhat people prefer in blind comparisons, and judged writing quality
| Benchmark | DeepSeek-V4.1-Flash | Rank among available | Best available |
|---|---|---|---|
| Arena (formerly LMArena)Arena Text: overall · reported | 1,474 | 27th of 153 | Gemini 4 Argon 1,525 |
- UGI Leaderboard: has not measured this model.
Sticking to the factsSummarising without inventing things, and factual answers
| Benchmark | DeepSeek-V4.1-Flash | Rank among available | Best available |
|---|---|---|---|
| Arena (formerly LMArena)Arena Text factuality: overall · reported | 1,467 | 24th of 109 | Gemini 4 Argon 1,511 |
- Vectara Hallucination Leaderboard: has not measured this model.
- SimpleQA Verified (Epoch AI): has not measured this model.
Speed under pressureGood decisions against a real clock (fast chess)
- BulletBench: has not measured this model.
Nearest alternatives
- Cheaper at similar intelligenceMiMo-V2.6-Flash$0.18 blended · Intelligence Index 37.9Compare with DeepSeek-V4.1-Flash →
- DeepSeek's newer model in this lineDeepSeek-V4.1-Flash-1017$0.52 blended · Intelligence Index 45.1Compare with DeepSeek-V4.1-Flash →
Against DeepSeek-V4.1-Flash at $0.52 blended, Intelligence Index 39.5. "Similar intelligence" means within 4 points or better.
Where to run it
34 hosts on OpenRouter at the standard service tier, with a median of $0.34 blended per million tokens; quantisation where the host states it: FP8, FP4, MXFP4. Cheapest first.
| Host | Quantisation | Input | Output | Blended |
|---|---|---|---|---|
| Morph | FP8 | $0.02 | $0.40 | $0.12 |
| Relace | FP4 | $0.01 | $0.50 | $0.13 |
| Wafer | MXFP4 | $0.05 | $0.57 | $0.18 |
| Alibaba | not stated | $0.10 | $0.42 | $0.18 |
| DeepInfra | FP8 | $0.14 | $0.42 | $0.21 |
| InferenceNet | FP8 | $0.10 | $0.60 | $0.23 |
| InferenceNet | FP8 | $0.10 | $0.60 | $0.23 |
| AtlasCloud | FP8 | $0.14 | $0.54 | $0.24 |
Other service tiers, not counted above: Parasail fast $0.45 in, $1.80 out; DeepInfra flex $0.11 in, $0.34 out; DeepInfra priority $0.21 in, $0.63 out. Flex and batch are cheaper but slower or queued; priority and fast cost more.
Prices per million tokens from OpenRouter's public endpoint list, checked 10 Nov 2026; the 26 dearer hosts are left out.
Against the alternatives
The models you would most likely weigh it against: the leaders of the groups above.
Scroll sideways to see every alternative.
| DeepSeek-V4.1-Flash | MiMo-V2.6-Pro most intelligent at $0.25–1 per million tokens | MiMo-V2.6-Flash cheapest with similar intelligence | Claude Opus 5.5 most intelligent available today | Claude Fable 5.2 highest-scoring overall available today | |
|---|---|---|---|---|---|
| Price per million tokens | $0.52 | $0.54 | $0.18 | $8.00 | $16.00 |
| Tokens a second | 369 | 43 | 46 | 96 | – |
| Intelligence Index | 39.5 | 46.3 | 37.9 | 57.6 | – |
| Spring Prompt overall | 51 | – | – | 82 | 90 |
| Rank among available models, by use case | |||||
| Product listings | 15th of 32 | – | – | 13th | 5th |
| Decks from an analysis | – | – | – | 7th | 4th |
| User surveys | – | – | – | 4th | 1st |
| Marketing planning | – | – | – | 9th | 1st |
| Agents and tool use | 31st of 159 | 23rd | 42nd | 3rd | – |
| Professional work | – | – | – | 3rd | – |
| Reasoning and knowledge | 42nd of 187 | 13th | 55th | 1st | – |
| Writing | 29th of 169 | 23rd | 56th | 5th | – |
| Sticking to the facts | 23rd of 139 | 10th | 46th | 3rd | – |
| Speed under pressure | – | – | – | – | 8th |
| Side by side → | Side by side → | Side by side → | |||
Shaded figures are better than DeepSeek-V4.1-Flash's. A dash means no result.
Research and examples
- DeepSeek-V4.1-Flash-1017 review: better at analysis, much worse at sticking to the product data
- We Told 11 AI Models What Happened in 2026. Most of Them Thought We Were Making It Up.
- Claude Haiku 5.5 Is Far Better Than Haiku 4.5 at a Tenth of the Price. GPT-6 Luna Still Beats It at Business Work.
In the playground: our pick in 0 of 20 tests, 86% of checks passed →
- Holographic trading cards (5 of 5 checks)
- Ball bouncing in a spinning hexagon (3 of 3 checks)
- The solar system, to scale in time (3 of 3 checks)
Where it has been measured
Fastest of the 23 models from developers based in China, and highest-scoring overall of the 5 models priced $0.25–1 per million tokens. 4th most intelligent of the 61 models priced $0.25–1 per million tokens. Among models available today, in the top quarter (by percentile) for 4 of the 5 use cases measured, led by writing, and agents and tool use.
- CatalogBench31 results
- Arena (formerly LMArena)11 results
- Artificial Analysis14 results
- Berkeley Function Calling Leaderboard (BFCL) V4Pending: latest results 16 Dec 2025
- Microsoft STATE-BenchPending: latest results 1 Jul 2026
- OpenHands IndexPending: latest results 30 Jun 2026
- tau2-benchPending: latest results 3 Aug 2026
- BulletBenchNot measured
- DeckBenchNot measured
- ROASBenchNot measured
- SurveyBenchNot measured
- APEX-AgentsNot measured
- GDP.pdfNot measured
- Remote Labor IndexNot measured
- SimpleQA Verified (Epoch AI)Not measured
- UGI LeaderboardNot measured
- Vectara Hallucination LeaderboardNot measured
- Vending-Bench 2Not measured
Every result
56 published results from 3 sources, each in the source's own units, with the configuration that produced it.
CatalogBench · measured by Spring Prompt · 31 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| CatalogBench (sales brief): channel rules broken% of products, lower is better | 1.8% | 30 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): claims to check% of products, lower is better | 0.0% | 1 of 32 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): content quality% of checks, higher is better | 96.2% | 23 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): failed outputs% of products, lower is better | 0.6% | 26 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): missing UK information% of products, lower is better | 0.0% | 1 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): not findable% of products, lower is better | 3.6% | 16 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): publish-ready listings% of products, higher is better | 47.0% | 12 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): reliably publish-ready% of products, higher is better | 23.2% | 11 of 34 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): unsupported claims% of products, lower is better | 39.9% | 14 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): unsupported claimsclaims per product, lower is better | 1.25 | 13 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): wrong attributes% of products, lower is better | 20.2% | 23 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench (sales brief): wrong category or variant% of products, lower is better | 2.4% | 18 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: channel compliance% of products, higher is better | 99.4% | 26 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: channel rules broken% of products, lower is better | 3.0% | 33 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: claims to check% of products, lower is better | 0.0% | 1 of 32 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: conflicts caught% of conflicts, higher is better | 100.0% | 1 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: content quality% of checks, higher is better | 96.7% | 20 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: cost per productUS dollars, lower is better | $0.0061 | 12 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: decision accuracy% of decisions, higher is better | 96.9% | 15 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: failed outputs% of products, lower is better | 0.0% | 1 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: field accuracy% of missing fields, higher is better | 94.4% | 8 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: invented values% of filled values, lower is better | 3.1% | 28 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: missing UK information% of products, lower is better | 0.0% | 1 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: not findable% of products, lower is better | 3.6% | 12 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: publish-ready listings% of products, higher is better | 63.7% | 16 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: reliably publish-ready% of products, higher is better | 41.1% | 17 of 34 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: time per productseconds, lower is better | 17.6 | 18 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: unsupported claims% of products, lower is better | 13.1% | 14 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: unsupported claimsclaims per product, lower is better | 0.54 | 18 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: wrong attributes% of products, lower is better | 20.2% | 26 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
| CatalogBench: wrong category or variant% of products, lower is better | 3.0% | 24 of 35 | DeepSeek-V4.1-Flash provider default reasoning |
30 Sep 2026 |
Not shown: Not your catalogue: invented products with generated images, so a model's result on your own feed can differ.
Arena (formerly LMArena) · reported by Arena (formerly LMArena) · 11 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Arena Agent: confirmed task successIPS effect estimate, higher is better | 0.08 0.07–0.09 |
8 of 50 | DeepSeek-V4.1-Flash (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Agent: praise over complaintIPS effect estimate, higher is better | 0.02 0.01–0.04 |
25 of 50 | DeepSeek-V4.1-Flash (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Agent: steerabilityIPS effect estimate, higher is better | 0.01 0.01–0.02 |
26 of 50 | DeepSeek-V4.1-Flash (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Agent: tool groundingIPS effect estimate, higher is better | 0.00 0.00–0.00 |
17 of 50 | DeepSeek-V4.1-Flash (max reasoning, arena agent) | 2 Oct 2026 |
| Arena Text factuality: overallArena rating, higher is better | 1,467 1,461–1,473 |
33 of 182 | DeepSeek-V4.1-Flash (max reasoning) | 2 Oct 2026 |
| Arena Text: business, management and financeArena rating, higher is better | 1,479 1,464–1,494 |
27 of 406 | DeepSeek-V4.1-Flash (max reasoning) | 2 Oct 2026 |
| Arena Text: creative writingArena rating, higher is better | 1,440 1,425–1,455 |
60 of 411 | DeepSeek-V4.1-Flash (max reasoning) | 2 Oct 2026 |
| Arena Text: expert promptsArena rating, higher is better | 1,517 1,498–1,535 |
21 of 363 | DeepSeek-V4.1-Flash (max reasoning) | 2 Oct 2026 |
| Arena Text: instruction followingArena rating, higher is better | 1,479 1,469–1,490 |
21 of 413 | DeepSeek-V4.1-Flash (max reasoning) | 2 Oct 2026 |
| Arena Text: overallArena rating, higher is better | 1,474 1,468–1,481 |
38 of 413 | DeepSeek-V4.1-Flash (max reasoning) | 2 Oct 2026 |
| Arena Text: writing, literature and languageArena rating, higher is better | 1,454 1,442–1,467 |
45 of 412 | DeepSeek-V4.1-Flash (max reasoning) | 2 Oct 2026 |
Not shown: Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.
Artificial Analysis · reported by Artificial Analysis · 14 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better | 39.5 | 66 of 413 | DeepSeek-V4.1-Flash (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better | 24.7 | 159 of 413 | DeepSeek-V4.1-Flash (no reasoning) | 20 Oct 2026 |
| Artificial Analysis: Humanity's Last Exam% of questions, higher is better | 39.2% | 93 of 419 | DeepSeek-V4.1-Flash (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Humanity's Last Exam% of questions, higher is better | 10.8% | 271 of 419 | DeepSeek-V4.1-Flash (no reasoning) | 20 Oct 2026 |
| Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better | 84.0% | 11 of 407 | DeepSeek-V4.1-Flash (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better | 55.3% | 262 of 407 | DeepSeek-V4.1-Flash (no reasoning) | 20 Oct 2026 |
| Artificial Analysis: Output speedtokens per second, higher is better | 231 | 17 of 137 | DeepSeek-V4.1-Flash (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Output speedtokens per second, higher is better | 369 | 3 of 137 | DeepSeek-V4.1-Flash (no reasoning) | 20 Oct 2026 |
| Artificial Analysis: SciCode% of problems, higher is better | 51.9% | 90 of 205 | DeepSeek-V4.1-Flash (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: SciCode% of problems, higher is better | 35.5% | 184 of 205 | DeepSeek-V4.1-Flash (no reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better | 26.8% | 53 of 204 | DeepSeek-V4.1-Flash (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better | 5.6% | 108 of 204 | DeepSeek-V4.1-Flash (no reasoning) | 20 Oct 2026 |
| Artificial Analysis: Time to first answer tokenseconds, lower is better | 9.37 | 59 of 137 | DeepSeek-V4.1-Flash (max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Time to first answer tokenseconds, lower is better | 0.63 | 13 of 137 | DeepSeek-V4.1-Flash (no reasoning) | 20 Oct 2026 |
Not shown: Not business work, and a blend: read the parts for any one task.
Embed a score badge
A badge for each of our benchmarks that ranked DeepSeek-V4.1-Flash, from the current results. Paste the code into a page or README; it links back to the benchmark.
CatalogBench · Reliably publish-ready, factual brief
<a href="https://springprompt.com/benchmarks/catalogbench?focus=deepseek-deepseek-v4-1-flash"><img src="https://springprompt.com/badge/catalogbench/deepseek-deepseek-v4-1-flash.svg" alt="DeepSeek-V4.1-Flash is #17 of 32 on CatalogBench (Reliably publish-ready, factual brief: 41.1%), measured by Spring Prompt" height="20"></a>
Will DeepSeek-V4.1-Flash work for your product?
Benchmarks aren't your task. Paste the prompt your product uses and run it beside the models you could use instead, with every answer checked and costed.
New here? Your email starts a free trial. No credit card needed.