Models / DeepSeek-V4-Pro (0423)
DeepSeek-V4-Pro (0423)
DeepSeek-V4-Pro (0423) is best for writing; 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 $1.77.
Superseded. DeepSeek-V4-Pro (0813) is DeepSeek's newer model in this line, at $1.98 blended: see DeepSeek-V4-Pro (0813) or compare the two.
Results 16 Nov 2026 · 91 results · 10 sources
Test DeepSeek-V4-Pro (0423) on your prompt → It runs beside the models you could use instead, with every answer checked and costed.
- Price per million tokens
- $1.42 in · $2.83 out
Typical host on OpenRouter, 10 Nov 2026 · compare prices - Speed
- Not measured
- Intelligence Index
- 30.4
Artificial Analysis, at max reasoning; 20.8–30.4 across 3 settings - Spring Prompt overall
- 31st of 36 score 28 of 100 · how it works
- Developer
- DeepSeek, based in China
Weights published - Released
- 24 Apr 2026
1,048,576 tokens of context
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 | 50th of 185Claude Opus 5.5 | 30th of 35Claude Fable 5.2 | $1.77 blended, 66th percentilecheapest: Mistral Nemo | – |
| Models priced $1–3 per million tokens | 10th of 34Muse Spark 1.3 | 5th of 8GPT-6.1 Terra | sets this group | – |
| Models with similar intelligence | sets this group | – | $1.77 blended, 43rd percentilecheapest: DeepSeek-V4-Flash (0731) | – |
| Models from developers based in China | 16th of 75MiMo-V2.6-Pro | 10th of 11Kimi K3.1 Pro | $1.77 blended, 86th percentilecheapest: Qwen3.5-9B | – |
| Open-weights models | 17th of 104MiMo-V2.6-Pro | 7th of 8Kimi K3.1 Pro | $1.77 blended, 90th percentilecheapest: Mistral Nemo | – |
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 |
|---|---|---|
| Deck | DeckBench | ≈ $0.046 |
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.
- DeckBench · cost per presentable deck
DeepSeek-V4-Pro (0423): no presentable decks (0 of 6 tasks). Cheapest: GPT-6.1 Terra at $0.13 (3 of 6 tasks).
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. With a handful of tasks, a pass or two either way is within chance. 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 $1–3 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
- CatalogBench: has not measured this model.
Decks from an analysisTurning a finished analysis into a deck you could present as it is
| Benchmark | DeepSeek-V4-Pro (0423) | Rank among available | Best available |
|---|---|---|---|
| DeckBenchDeckBench: deck rating | 542 | 32nd of 32 | Qwen4-Max 1,585 |
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
| Benchmark | DeepSeek-V4-Pro (0423) | Rank among available | Best available |
|---|---|---|---|
| ROASBenchROASBench: overall score | 28.3 | 26th of 32 | Claude Fable 5.2 61.3 |
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls
| Benchmark | DeepSeek-V4-Pro (0423) | Rank among available | Best available |
|---|---|---|---|
| OpenHands IndexOpenHands Index: average · reported | 40.7% | 27th of 31 | Claude Fable 5 81.0% |
| Microsoft STATE-BenchSTATE-Bench v0.7: task success (pass@1) · reported | 47.2% | 4th of 4 | GPT-5.4 55.7% |
| Artificial AnalysisArtificial Analysis: Terminal-Bench 4.0 · reported | 14.6% | 30th of 107 | Kimi K3.1 Pro 67.2% |
| Artificial AnalysisArtificial Analysis: Terminal-Bench 2.1 · reported | 64.8% | 40th of 107 | Claude Fable 5.1 91.4% |
| Artificial AnalysisArtificial Analysis: τ-bench banking · reported | 30.1% | 33rd of 104 | Grok 4.6 50.7% |
| Vending-Bench 2Vending-Bench 2: money after a year · reported | $3,285 | 36th of 53 | GPT-6 Astra $15,515 |
- tau2-bench: has not measured this model.
- Berkeley Function Calling Leaderboard (BFCL) V4: pending. Its latest results are dated 16 Dec 2025, before this model was released.
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-Pro (0423) | Rank among available | Best available |
|---|---|---|---|
| Artificial AnalysisArtificial Analysis: Artificial Analysis Intelligence Index · reported | 30.4 | 50th of 185 | Claude Opus 5.5 57.6 |
| Artificial AnalysisArtificial Analysis: Humanity's Last Exam · reported | 37.5% | 53rd of 186 | Claude Opus 5.5 61.4% |
| Artificial AnalysisArtificial Analysis: GPQA Diamond · reported | 90.5% | 34th of 168 | GPT-6 Astra 96.3% |
WritingWhat people prefer in blind comparisons, and judged writing quality
| Benchmark | DeepSeek-V4-Pro (0423) | Rank among available | Best available |
|---|---|---|---|
| Arena (formerly LMArena)Arena Text: overall · reported | 1,458 | 40th of 153 | Gemini 4 Argon 1,525 |
| UGI LeaderboardUGI: writing score · reported | 68.4 | 14th of 124 | Gemini 3.8 Flash 78.6 |
Sticking to the factsSummarising without inventing things, and factual answers
| Benchmark | DeepSeek-V4-Pro (0423) | Rank among available | Best available |
|---|---|---|---|
| Vectara Hallucination LeaderboardVectara: hallucination rate · reported | 8.6% | 31st of 71 | GPT-5.4 nano 3.1% |
| Arena (formerly LMArena)Arena Text factuality: overall · reported | 1,455 | 42nd of 109 | Gemini 4 Argon 1,511 |
| SimpleQA Verified (Epoch AI)SimpleQA Verified: correct answers · reported | 47.0% | 30th of 68 | GPT-6 Astra 75.6% |
Speed under pressureGood decisions against a real clock (fast chess)
- BulletBench: too slow for Blitz 3+2: lost its first 4 games on time, at a median 7.3 s a move; too slow for Bullet 60s: lost its first 4 games on time, at a median 5.0 s a move; too slow for Lightning 10+1: lost its first 4 games on time, at a median 6.2 s a move.
Nearest alternatives
- Cheaper at similar intelligenceMiMo-V2.6-Flash$0.18 blended · Intelligence Index 37.9Compare with DeepSeek-V4-Pro (0423) →
- DeepSeek's newer model in this lineDeepSeek-V4-Pro (0813)$1.98 blended · Intelligence Index 36.0Compare with DeepSeek-V4-Pro (0423) →
Against DeepSeek-V4-Pro (0423) at $1.77 blended, Intelligence Index 30.4. "Similar intelligence" means within 4 points or better.
Where to run it
13 hosts on OpenRouter at the standard service tier, with a median of $1.77 blended per million tokens; quantisation where the host states it: FP4, FP8. Cheapest first.
| Host | Quantisation | Input | Output | Blended |
|---|---|---|---|---|
| Relace | FP4 | $0.90 | $1.50 | $1.05 |
| SiliconFlow | FP8 | $1.00 | $1.50 | $1.12 |
| StreamLake | FP8 | $0.96 | $1.91 | $1.19 |
| GMICloud | FP8 | $0.96 | $1.91 | $1.20 |
| DigitalOcean | not stated | $1.04 | $2.09 | $1.31 |
| DeepInfra | FP8 | $1.30 | $2.60 | $1.62 |
| Alibaba | FP8 | $1.42 | $2.83 | $1.77 |
| Alibaba | FP8 | $1.56 | $3.12 | $1.95 |
Prices per million tokens from OpenRouter's public endpoint list, checked 10 Nov 2026; the 5 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-Pro (0423) | Muse Spark 1.3 most intelligent at $1–3 per million tokens | DeepSeek-V4-Flash (0731) cheapest with similar intelligence | Claude Opus 5.5 most intelligent available today | MiMo-V2.6-Pro most intelligent from China | |
|---|---|---|---|---|---|
| Price per million tokens | $1.77 | $2.00 | $0.20 | $8.00 | $0.54 |
| Tokens a second | – | 254 | – | 96 | 43 |
| Intelligence Index | 30.4 | 48.1 | 34.3 | 57.6 | 46.3 |
| Spring Prompt overall | 28 | 59 | – | 82 | – |
| Rank among available models, by use case | |||||
| Product listings | – | 17th | – | 13th | – |
| Decks from an analysis | 32nd of 32 | 25th | – | 7th | – |
| User surveys | – | – | – | 4th | – |
| Marketing planning | 26th of 32 | 25th | – | 9th | – |
| Agents and tool use | 92nd of 159 | 13th | 43rd | 3rd | 23rd |
| Professional work | – | 8th | – | 3rd | – |
| Reasoning and knowledge | 51st of 187 | 11th | 47th | 1st | 13th |
| Writing | 30th of 169 | 7th | – | 5th | 23rd |
| Sticking to the facts | 56th of 139 | 11th | 105th | 3rd | 10th |
| Speed under pressure | 36th of 36 | 33rd | – | – | – |
| Side by side → | Side by side → | Side by side → | Side by side → | ||
Shaded figures are better than DeepSeek-V4-Pro (0423)'s. A dash means no result.
Where it has been measured
10th most intelligent of the 34 models priced $1–3 per million tokens. Among models available today, in the top quarter (by percentile) for writing; in the bottom quarter for decks from an analysis, marketing planning and speed under pressure.
Artificial Analysis lists it but has not published a speed for it yet
- DeckBench17 results
- ROASBench9 results
- Arena (formerly LMArena)14 results
- Artificial Analysis29 results
- Microsoft STATE-Bench6 results
- OpenHands Index6 results
- SimpleQA Verified (Epoch AI)1 result
- UGI Leaderboard6 results
- Vectara Hallucination Leaderboard2 results
- Vending-Bench 21 result
- Berkeley Function Calling Leaderboard (BFCL) V4Pending: latest results 16 Dec 2025
- BulletBenchNot measured
- CatalogBenchNot measured
- SurveyBenchNot measured
- APEX-AgentsNot measured
- GDP.pdfNot measured
- Remote Labor IndexNot measured
- tau2-benchNot measured
Every result
91 published results from 10 sources, each in the source's own units, with the configuration that produced it.
DeckBench · measured by Spring Prompt · 17 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| DeckBench: accurate decks% of tasks, higher is better | 0.0% | 29 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: caveat dropped% of tasks, lower is better | 16.7% | 15 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: clean layout% of tasks, higher is better | 33.3% | 16 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: cost per deckUS dollars, lower is better | $0.0465 | 15 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: deck ratingrating, higher is better | 542 408–625 |
34 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: design quality% of the maximum, higher is better | 40.9% | 32 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: draft figure quoted% of tasks, lower is better | 33.3% | 19 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: findings missing% of tasks, lower is better | 16.7% | 23 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: head-to-head win rate% of comparisons, higher is better | 24.4% | 34 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: layout defects% of tasks, lower is better | 66.7% | 16 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: misleading metric used% of tasks, lower is better | 0.0% | 1 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: presentable decks% of tasks, higher is better | 0.0% | 15 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: recommendation late or wrong% of tasks, lower is better | 66.7% | 29 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: slides needing work% of tasks, lower is better | 83.3% | 31 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: time per deckseconds, lower is better | 332 | 29 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: unsupported claims% of tasks, lower is better | 66.7% | 33 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
| DeckBench: unsupported numbers% of tasks, lower is better | 0.0% | 1 of 34 | DeepSeek-V4-Pro (0423) provider default reasoning |
6 Oct 2026 |
Not shown: Not your analysis or your brand: invented companies, six public tasks, one deck per model per task at the provider's default reasoning setting.
ROASBench · measured by Spring Prompt · 9 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| ROASBench: audience scorescore out of 100, higher is better | 44.8 | 31 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: business scorescore out of 100, higher is better | 23.8 | 27 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: consistency scorescore out of 100, higher is better | 9.1 | 33 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: contribution profitsimulated US dollars, higher is better | $301,289 | 26 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: cost of a runUS dollars, lower is better | $0.20 | 19 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: months over budgetmonths of 12, lower is better | 0 | 1 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: overall scorescore out of 100, higher is better | 28.3 | 27 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: planning scorescore out of 100, higher is better | 54.0 | 32 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
| ROASBench: return on ad spendprofit per $1 spent, higher is better | $0.26 | 26 of 35 | DeepSeek-V4-Pro (0423) provider default reasoning |
29 Sep 2026 |
Not shown: Not real ad performance: the market is a deterministic simulation of one invented skincare brand.
Arena (formerly LMArena) · reported by Arena (formerly LMArena) · 14 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Arena Text factuality: overallArena rating, higher is better | 1,455 1,452–1,458 |
57 of 182 | DeepSeek-V4-Pro (0423) | 2 Oct 2026 |
| Arena Text factuality: overallArena rating, higher is better | 1,448 1,444–1,451 |
72 of 182 | DeepSeek-V4-Pro (0423, high reasoning) | 2 Oct 2026 |
| Arena Text: business, management and financeArena rating, higher is better | 1,460 1,453–1,467 |
60 of 406 | DeepSeek-V4-Pro (0423) | 2 Oct 2026 |
| Arena Text: business, management and financeArena rating, higher is better | 1,448 1,441–1,455 |
79 of 406 | DeepSeek-V4-Pro (0423, high reasoning) | 2 Oct 2026 |
| Arena Text: creative writingArena rating, higher is better | 1,446 1,439–1,453 |
48 of 411 | DeepSeek-V4-Pro (0423) | 2 Oct 2026 |
| Arena Text: creative writingArena rating, higher is better | 1,441 1,433–1,448 |
59 of 411 | DeepSeek-V4-Pro (0423, high reasoning) | 2 Oct 2026 |
| Arena Text: expert promptsArena rating, higher is better | 1,480 1,471–1,488 |
73 of 363 | DeepSeek-V4-Pro (0423) | 2 Oct 2026 |
| Arena Text: expert promptsArena rating, higher is better | 1,476 1,467–1,484 |
79 of 363 | DeepSeek-V4-Pro (0423, high reasoning) | 2 Oct 2026 |
| Arena Text: instruction followingArena rating, higher is better | 1,452 1,447–1,458 |
62 of 413 | DeepSeek-V4-Pro (0423) | 2 Oct 2026 |
| Arena Text: instruction followingArena rating, higher is better | 1,448 1,442–1,454 |
67 of 413 | DeepSeek-V4-Pro (0423, high reasoning) | 2 Oct 2026 |
| Arena Text: overallArena rating, higher is better | 1,458 1,454–1,462 |
62 of 413 | DeepSeek-V4-Pro (0423) | 2 Oct 2026 |
| Arena Text: overallArena rating, higher is better | 1,455 1,451–1,459 |
72 of 413 | DeepSeek-V4-Pro (0423, high reasoning) | 2 Oct 2026 |
| Arena Text: writing, literature and languageArena rating, higher is better | 1,451 1,444–1,457 |
53 of 412 | DeepSeek-V4-Pro (0423) | 2 Oct 2026 |
| Arena Text: writing, literature and languageArena rating, higher is better | 1,443 1,437–1,449 |
67 of 412 | DeepSeek-V4-Pro (0423, high reasoning) | 2 Oct 2026 |
Not shown: Not an error rate: claims that cannot be checked on the web are skipped, and preference still carries most of the weight.
Artificial Analysis · reported by Artificial Analysis · 29 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Artificial Analysis: Artificial Analysis Coding Indexindex score, higher is better | 58.7 | 77 of 187 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Coding Indexindex score, higher is better | 59.4 | 75 of 187 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better | 30.1 | 115 of 413 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better | 30.4 | 112 of 413 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better | 20.8 | 200 of 413 | DeepSeek-V4-Pro (0423, no reasoning) | 20 Oct 2026 |
| Artificial Analysis: GPQA Diamond% of questions, higher is better | 90.5% | 60 of 360 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: GPQA Diamond% of questions, higher is better | 88.8% | 86 of 360 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: GPQA Diamond% of questions, higher is better | 71.7% | 239 of 360 | DeepSeek-V4-Pro (0423, no reasoning) | 20 Oct 2026 |
| Artificial Analysis: Humanity's Last Exam% of questions, higher is better | 35.2% | 118 of 419 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: Humanity's Last Exam% of questions, higher is better | 37.5% | 104 of 419 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Humanity's Last Exam% of questions, higher is better | 8.2% | 297 of 419 | DeepSeek-V4-Pro (0423, no reasoning) | 20 Oct 2026 |
| Artificial Analysis: IFBench% of instructions, higher is better | 71.3% | 52 of 288 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: IFBench% of instructions, higher is better | 76.5% | 17 of 288 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: IFBench% of instructions, higher is better | 45.8% | 161 of 288 | DeepSeek-V4-Pro (0423, no reasoning) | 20 Oct 2026 |
| Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better | 70.3% | 195 of 407 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better | 74.7% | 154 of 407 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better | 53.0% | 269 of 407 | DeepSeek-V4-Pro (0423, no reasoning) | 20 Oct 2026 |
| Artificial Analysis: SciCode% of problems, higher is better | 50.8% | 104 of 205 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 2.1% of tasks, higher is better | 64.8% | 79 of 188 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 2.1% of tasks, higher is better | 64.0% | 81 of 188 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better | 14.6% | 78 of 204 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench Hard% of tasks, higher is better | 41.7% | 48 of 282 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench Hard% of tasks, higher is better | 46.2% | 29 of 282 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: Terminal-Bench Hard% of tasks, higher is better | 36.4% | 69 of 282 | DeepSeek-V4-Pro (0423, no reasoning) | 20 Oct 2026 |
| Artificial Analysis: τ-bench banking% of tasks, higher is better | 26.2% | 71 of 175 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: τ-bench banking% of tasks, higher is better | 30.1% | 61 of 175 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: τ²-bench telecom% of tasks, higher is better | 94.2% | 30 of 286 | DeepSeek-V4-Pro (0423, high reasoning) | 20 Oct 2026 |
| Artificial Analysis: τ²-bench telecom% of tasks, higher is better | 96.2% | 13 of 286 | DeepSeek-V4-Pro (0423, max reasoning) | 20 Oct 2026 |
| Artificial Analysis: τ²-bench telecom% of tasks, higher is better | 91.2% | 52 of 286 | DeepSeek-V4-Pro (0423, no reasoning) | 20 Oct 2026 |
Not shown: Not your codebase or tools.
Microsoft STATE-Bench · reported by Microsoft STATE-Bench · 6 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| STATE-Bench v0.7: consistency (pass^5)% of tasks, higher is better | 25.3% | 5 of 5 | DeepSeek-V4-Pro (0423, statebench) | 25 May 2026 |
| STATE-Bench v0.7: customer support task success (pass@1)% of tasks, higher is better | 45.6% | 4 of 5 | DeepSeek-V4-Pro (0423, statebench) | 25 May 2026 |
| STATE-Bench v0.7: shopping assistant task success (pass@1)% of tasks, higher is better | 48.4% | 4 of 5 | DeepSeek-V4-Pro (0423, statebench) | 25 May 2026 |
| STATE-Bench v0.7: task success (pass@1)% of tasks, higher is better | 47.2% | 4 of 5 | DeepSeek-V4-Pro (0423, statebench) | 25 May 2026 |
| STATE-Bench v0.7: travel task success (pass@1)% of tasks, higher is better | 47.6% | 4 of 5 | DeepSeek-V4-Pro (0423, statebench) | 25 May 2026 |
| STATE-Bench v0.7: user experiencescore from 1 to 5, higher is better | 3.36 | 5 of 5 | DeepSeek-V4-Pro (0423, statebench) | 25 May 2026 |
Not shown: Not the model alone: it runs inside STATE-Bench's agent loop against simulated users, and versions of the benchmark are not comparable.
OpenHands Index · reported by OpenHands Index · 6 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| OpenHands Index: average% resolved, higher is better | 40.7% | 29 of 34 | DeepSeek-V4-Pro (0423, openhands) | 30 Jun 2026 |
| OpenHands Index: front end (SWE-Bench Multimodal)% resolved, higher is better | 36.8% | 11 of 34 | DeepSeek-V4-Pro (0423, openhands) | 30 Jun 2026 |
| OpenHands Index: greenfield (Commit0)% resolved, higher is better | 12.5% | 25 of 34 | DeepSeek-V4-Pro (0423, openhands) | 30 Jun 2026 |
| OpenHands Index: information gathering (GAIA)% resolved, higher is better | 12.7% | 33 of 34 | DeepSeek-V4-Pro (0423, openhands) | 30 Jun 2026 |
| OpenHands Index: issue resolution (SWE-Bench)% resolved, higher is better | 73.2% | 22 of 34 | DeepSeek-V4-Pro (0423, openhands) | 30 Jun 2026 |
| OpenHands Index: testing (SWT-Bench)% resolved, higher is better | 68.1% | 18 of 34 | DeepSeek-V4-Pro (0423, openhands) | 30 Jun 2026 |
Not shown: Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
SimpleQA Verified (Epoch AI) · reported by SimpleQA Verified (Epoch AI) · 1 result
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| SimpleQA Verified: correct answers% of questions, higher is better | 47.0% | 33 of 85 | DeepSeek-V4-Pro (0423, max reasoning) | 27 Aug 2026 |
Not shown: Not answers grounded in your documents; tests what the model remembers.
UGI Leaderboard · reported by UGI Leaderboard · 6 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| UGI: requested-length error% off the requested word count, lower is better | 42.0% | 304 of 379 | DeepSeek-V4-Pro (0423, no reasoning) | 26 Apr 2026 |
| UGI: requested-length error% off the requested word count, lower is better | 34.0% | 278 of 379 | DeepSeek-V4-Pro (0423, thinking reasoning) | 26 Apr 2026 |
| UGI: style adherencescore from 0 to 1, higher is better | 0.35 | 182 of 379 | DeepSeek-V4-Pro (0423, no reasoning) | 26 Apr 2026 |
| UGI: style adherencescore from 0 to 1, higher is better | 0.36 | 149 of 379 | DeepSeek-V4-Pro (0423, thinking reasoning) | 26 Apr 2026 |
| UGI: writing scorescore out of 100, higher is better | 60.3 | 102 of 379 | DeepSeek-V4-Pro (0423, no reasoning) | 26 Apr 2026 |
| UGI: writing scorescore out of 100, higher is better | 68.4 | 36 of 379 | DeepSeek-V4-Pro (0423, thinking reasoning) | 26 Apr 2026 |
Not shown: Not other format limits such as character counts or bullet counts.
Vectara Hallucination Leaderboard · reported by Vectara Hallucination Leaderboard · 2 results
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Vectara: answer rate% of documents, higher is better | 97.2% | 87 of 108 | DeepSeek-V4-Pro (0423) | 22 Sep 2026 |
| Vectara: hallucination rate% of summaries, lower is better | 8.6% | 44 of 108 | DeepSeek-V4-Pro (0423) | 22 Sep 2026 |
Not shown: Not a quality score: a low rate usually means content filters were triggered, and hallucination rates are measured on answered documents only.
Vending-Bench 2 · reported by Vending-Bench 2 · 1 result
| Measure | Value | Rank | Configuration | Dated |
|---|---|---|---|---|
| Vending-Bench 2: money after a yearUS dollars, higher is better | $3,285 | 42 of 63 | DeepSeek-V4-Pro (0423) | 1 Oct 2026 |
Not shown: Not a real business; one simulated market with set rules.
Embed a score badge
A badge for each of our benchmarks that ranked DeepSeek-V4-Pro (0423), from the current results. Paste the code into a page or README; it links back to the benchmark.
DeckBench · Deck rating
<a href="https://springprompt.com/benchmarks/deckbench?focus=deepseek-deepseek-v4-pro"><img src="https://springprompt.com/badge/deckbench/deepseek-deepseek-v4-pro.svg" alt="DeepSeek-V4-Pro (0423) is #33 of 33 on DeckBench (Deck rating: 542), measured by Spring Prompt" height="20"></a>
ROASBench · Score
<a href="https://springprompt.com/benchmarks/roasbench?focus=deepseek-deepseek-v4-pro"><img src="https://springprompt.com/badge/roasbench/deepseek-deepseek-v4-pro.svg" alt="DeepSeek-V4-Pro (0423) is #26 of 33 on ROASBench (Score: 28.3), measured by Spring Prompt" height="20"></a>
Will DeepSeek-V4-Pro (0423) 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.