Who ranks #1 on the Coding leaderboard?
As of August 22, 2026, DeepSeek-V4-Pro-0813 by DeepSeek ranks #1 on SWE-bench Verified at 96.4%. API pricing is $1.32/M input and $3.96/M output.
Coding
🏆 LeaderboardAs of August 22, 2026, DeepSeek-V4-Pro-0813 is #1 for coding at 96.4%. Ranked by the SWE-bench Verified score Best AI for coding, ranked by SWE-bench Verified and Terminal-Bench. Compare Claude vs ChatGPT for coding and see API price on every row.
Updated August 22, 2026282 models33 providers
 DeepSeek-V4-Pro-0813OSS DeepSeek · Open Source | 96.4% | 87.5% | — | 1446 | 56.2% | — | 82.3% | — | 87.9% | — | 35.8% | — | — | 41.5% | 62.7% | 62.8% | — | — | — | $5.28 |
 GPT-5.6 Sol OpenAI · Proprietary | 96.2% | 82.6% | 56.9% | 1620.27 | — | — | 80.5% | 64.6% | 88.8% | — | 86.7% | — | — | 52.9% | 72.7% | 72.7% | 47.5% | 67.2% | — | $35.00 |
 Claude Opus 5 Anthropic · Proprietary | 96% | 89.0% | 55.7% | 1711.88 | — | — | 88.4% | 79.2% | 84.6% | — | 91.7% | — | — | 57.5% | 73.7% | 73.7% | 53.4% | 70% | — | $30.00 |
 Grok 4.6 xAI · Proprietary | 95.6% | 88.2% | — | 1631 | — | — | 76.2% | — | 78.3% | — | — | — | — | 44.6% | — | 67.5% | 61.3% | — | — | $8.00 |
Claude Mythos 5 Anthropic · Proprietary | 95.5% | — | — | — | — | 88% | — | 80.3% | 88% | — | — | — | — | — | — | — | — | — | 80.3% | $60.00 |
Claude Fable 5 Anthropic · Proprietary | 95% | 89.8% | 60.2% | 1627 | — | — | 90.3% | 80% | 84.3% | — | 72.3% | — | — | 55.1% | 69.9% | 69.9% | 46.3% | 72.9% | 80% | $60.00 |
Claude Mythos Preview Anthropic · Proprietary | 93.9% | — | — | — | 82% | — | — | 77.8% | — | — | — | — | — | — | — | — | — | — | — | $60.00 |
 Kimi K3OSS Moonshot AI · Open Source | 93.4% | 87.2% | 58.7% | 1681.75 | — | 88.3% | 85.0% | — | 88.3% | — | — | — | — | 16.1% | 67.5% | 68.5% | 44.2% | — | — | $18.00 |
GPT-5.6 Luna OpenAI · Proprietary | 93% | — | 52.5% | 1522.94 | — | — | 77.1% | 62.7% | 84.7% | — | 72.9% | — | — | 44.5% | 67.2% | 67.2% | 39.8% | — | — | $1.40 |
DeepSeek-V4-Flash-0731OSS DeepSeek · Open Source | 88.8% | 87.3% | 49.9% | — | — | — | 74.7% | — | 82.7% | — | — | — | — | 38.6% | 54.4% | 53.3% | — | — | — | $1.76 |
Claude Opus 4.8 Anthropic · Proprietary | 88.6% | 87.8% | 53.5% | 1539 | 70.0% | — | 82.7% | 69.2% | 71.9% | — | — | — | — | 47.3% | 58.2% | 59.0% | 46.5% | 63.8% | — | $30.00 |
Claude Opus 4.7 Anthropic · Proprietary | 87.6% | 85.1% | 54.5% | 1558 | 68.5% | 80.2% | 71% | 64.3% | 68.5% | — | 47.1% | — | — | 43.9% | 54.2% | — | 38.5% | 64.8% | — | $30.00 |
 Muse Spark 1.2 Meta · Proprietary | 86.6% | — | — | 1535 | — | — | 79.1% | — | 82.9% | — | 49.5% | — | — | 29.9% | — | 54.9% | — | — | — | $5.50 |
Grok 4.5 xAI · Proprietary | 86.6% | 87.3% | 54.0% | 1555 | — | 83.3% | 69% | 64.7% | 83.3% | — | — | — | — | 36.6% | 53% | 53.8% | 42.4% | — | — | $8.00 |
 Qwen3.8 MaxOSS Qwen · Open Source | 85.6% | 87.8% | — | 1667 | — | 86.6% | 64.7% | 67.7% | 86.6% | — | 73% | — | — | 24.0% | — | 57.5% | — | — | — | $8.00 |
Claude Sonnet 5 Anthropic · Proprietary | 85.2% | 82.4% | 53.6% | 1544.16 | 80.4% | 80.4% | 81.3% | 63.2% | 74.5% | — | — | — | — | 44.4% | 53.9% | 53.9% | 38.8% | 61.2% | — | $12.00 |
GPT-5.5 OpenAI · Proprietary | 82.6% | 85.3% | 56.1% | 1504.74 | 82.7% | 84.7% | 69.8% | 58.6% | 76.4% | — | — | — | — | 45.2% | 70.0% | 67.0% | 43% | 64.3% | 58.6% | $35.00 |
Muse Spark 1.1 Meta · Proprietary | 82% | 85.9% | 58.2% | 1538 | — | — | 72.2% | 61.5% | 80% | — | — | — | — | 31.1% | 53.3% | 53.3% | — | — | — | $5.50 |
Claude Opus 4.5 Anthropic · Proprietary | 80.9% | 75.0% | — | 1468 | 58.4% | 63.1% | — | — | — | — | 23.6% | — | — | — | — | — | — | — | — | $30.00 |
 Gemini 3.7 Flash Google · Proprietary | 80.8% | 88.7% | — | 1587 | — | — | 70.4% | — | 85.8% | — | — | — | — | 34.8% | — | 65.5% | — | — | — | $4.50 |
Claude Opus 4.6 Anthropic · Proprietary | 80.8% | — | — | 1537 | 65.4% | 79.8% | 57.6% | — | — | — | — | — | — | — | 27.6% | — | — | — | — | $30.00 |
Gemini 3.1 Pro Google · Proprietary | 80.6% | 88.5% | 59% | 1447 | 67.4% | 80.2% | 32.0% | 54.2% | 70.8% | — | — | — | — | 17.3% | 9.7% | 11.7% | — | — | — | $17.50 |
DeepSeek-V4-Pro-MaxOSS DeepSeek · Open Source | 80.6% | 93.5% | 50% | — | 67.9% | — | — | 55.4% | — | — | — | — | — | — | — | — | — | — | — | $5.22 |
 MiniMax M3OSS MiniMax · Open Source | 80.5% | 82.2% | 45.4% | 1490 | 46.1% | — | 47.6% | 59% | 66% | — | — | — | — | 19.9% | 20.4% | — | 14.7% | — | — | $1.50 |
Qwen3.7 Max Qwen · Proprietary | 80.4% | 87.1% | 53.5% | 1517 | 59.2% | — | 47.7% | 60.6% | 61.0% | — | 46.8% | 91.6% | — | 13.1% | 17.7% | — | — | — | — | $5.00 |
MiniMax M2.5OSS MiniMax · Open Source | 80.2% | 79.2% | — | 1384 | 41.6% | 42.7% | 14.8% | 55.4% | — | — | 6.7% | — | — | — | — | — | — | — | — | $1.50 |
Kimi K2.6OSS Moonshot AI · Open Source | 80.2% | 86.8% | 52.2% | 1509 | 57.3% | — | 37.9% | 58.6% | 53.6% | — | — | 89.6% | — | 27.8% | 23.9% | — | — | 47.6% | — | $4.93 |
 Inkling-SmallOSS Thinking Machines · Open Source · via Thinking Machines Lab | 80.2% | 85.9% | 48.7% | — | — | — | 19.1% | 55.9% | 64.7% | — | — | — | — | 13.7% | — | — | — | — | — | $1.50 |
GPT-5.2 OpenAI · Proprietary | 80% | 85.4% | — | 1418 | 51.7% | 64.9% | 53.5% | 79.9% | — | — | 54.8% | — | — | — | — | — | — | — | — | $15.75 |
Gemini 3.6 Flash Google · Proprietary | 79.6% | 88.1% | 52.7% | 1527.83 | — | 78% | 64.0% | 58.7% | 78% | — | — | — | — | 30.9% | 48.6% | 46.7% | — | — | 58.7% | $4.50 |
Claude Sonnet 4.6 Anthropic · Proprietary | 79.6% | 82.1% | 46.8% | 1524 | 59.5% | 53.4% | 51.5% | — | 57.3% | — | — | — | — | 39.9% | 31.8% | 29.9% | — | 49% | — | $18.00 |
Gemini 3.5 Flash Google · Proprietary | 79.3% | 87.6% | 53.1% | 1506.38 | 67.4% | 76.2% | 48.7% | 55.1% | 76.2% | — | — | — | — | 26.8% | 28.3% | 36.1% | — | 49.8% | — | $10.50 |
DeepSeek-V4-Flash-MaxOSS DeepSeek · Open Source | 79% | 91.6% | 44.9% | — | 56.9% | — | — | 52.6% | — | — | — | — | — | — | — | — | — | — | — | $0.42 |
 MiMo-V2.5-ProOSS Xiaomi · Open Source | 78.9% | 81.3% | 50.2% | 1474 | 68.4% | — | 34.1% | 57.2% | 57.3% | 75.6% | — | 39.6% | — | 21.6% | 19.5% | — | — | — | — | $1.30 |
Qwen3.6 Plus Qwen · Proprietary | 78.8% | 86.0% | 40.7% | 1460 | 44.9% | — | 25.6% | 56.6% | 53.2% | — | — | 87.1% | — | 11.1% | 2.6% | — | — | — | — | $3.50 |
 GLM-5.2OSS Z AI · Open Source | 78.7% | 69.5% | 50.5% | 1593.25 | — | — | 64.0% | 62.1% | 82.7% | — | — | — | — | 37.9% | 41.5% | 43.8% | 24.5% | 54.6% | — | $5.80 |
DeepSeek-V4-Flash-0423OSS DeepSeek · Open Source | 78.6% | 88.4% | — | — | 56.6% | — | — | 52.3% | — | — | — | — | — | — | — | — | — | — | — | $0.30 |
Kimi K2.7 CodeOSS Moonshot AI · Open Source | 78.2% | 82.0% | 47.5% | 1473 | — | — | 47.2% | — | 67.0% | — | — | — | — | 25.4% | 30.5% | 30.5% | 30.1% | — | — | $4.93 |
 Hy3OSS Tencent · Open Source | 78% | — | — | 1522 | — | — | — | 57.9% | 71.7% | — | — | — | — | — | 28% | — | — | — | — | $0.66 |
Gemini 3 Flash Google · Proprietary | 78% | 85.6% | — | 1438 | 51.7% | 64.3% | 20.2% | — | 53.9% | — | 39.1% | — | — | 6.4% | 5.1% | — | — | — | — | $3.50 |
GLM-5OSS Z AI · Open Source | 77.8% | — | — | 1436 | 56.2% | 52.4% | — | — | — | — | — | — | — | — | — | — | — | — | — | $4.20 |
Qwen3.7-Plus Qwen · Proprietary | 77.7% | — | 51.3% | — | 70.3% | — | 46.4% | 57.6% | 52.8% | — | — | 89.6% | — | 12.9% | — | — | 10.2% | — | — | $1.60 |
 Mistral Medium 3.5OSS Mistral · Open Source · via Mistral AI | 77.6% | — | 39.6% | 1265 | 30.3% | — | 2.9% | — | 39.0% | — | — | — | — | 5.1% | — | — | — | — | — | $9.00 |
Qwen3.6-27BOSS Qwen · Open Source | 77.2% | — | — | — | 44.9% | — | 11.9% | 53.5% | — | — | — | 83.9% | — | — | — | — | — | — | — | $4.20 |
GPT-5.4 OpenAI · Proprietary | 76.9% | 84.1% | 56.6% | 1390 | 62.2% | 81.8% | 67.4% | 57.7% | — | — | 67.8% | — | — | 35.0% | 55.5% | 51.8% | — | — | 88.4% | $17.50 |
Kimi K2.5OSS Moonshot AI · Open Source | 76.8% | — | 48.7% | 1430.56 | 50.8% | 43.2% | — | 50.7% | — | — | — | 85% | — | — | — | — | — | 31.9% | — | $3.68 |
 Seed 2.0 Pro ByteDance · Proprietary | 76.5% | — | — | — | — | — | — | — | — | — | — | 87.8% | — | — | — | — | — | — | — | $3.50 |
Qwen3.5-397B-A17BOSS Qwen · Open Source | 76.4% | — | — | 1400 | 52.5% | — | — | — | — | — | — | 83.6% | — | — | — | — | — | — | — | $4.20 |
GPT-5.1 Thinking OpenAI · Proprietary | 76.3% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | $11.25 |
GPT-5.1 Instant OpenAI · Proprietary | 76.3% | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | $11.25 |
Showing 1–50 of 282 models
Next step
You found the model. Now ship the product.
Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.
The index
All Large Language Models
282 models across 33 providers. Search or jump to a lab — every model page stays linked here.
OpenAI
51 modelsGPT-5.6 CyberGPT-5.6 LunaGPT-5.6 SolGPT-5.6 TerraGPT-5.5 InstantGPT-5.5GPT-5.5 ProGPT-5.4 miniGPT-5.4 nanoGPT-5.4GPT-5.3 ChatGPT-5.3 CodexGPT-5.2 CodexGPT-5.2GPT-5.2 ProGPT-5.1 CodexGPT-5.1GPT-5.1 Codex HighGPT-5.1 Codex MiniGPT-5.1 HighGPT-5.1 InstantGPT-5.1 MediumGPT-5.1 ThinkingGPT-5 CodexGPT-5GPT-5 HighGPT-5 MediumGPT-5 miniGPT-5 nanoGPT OSS 120BGPT OSS 120B HighGPT OSS 20BGPT OSS 20B Higho3-proo3o4-miniGPT-4.1GPT-4.1 miniGPT-4.1 nanoo3-minio1o1-proo1-minio1-previewGPT-4oGPT-4o miniChatGPT-4o LatestGPT-4oGPT-4 TurboGPT-4GPT-3.5 Turbo
Anthropic
22 modelsClaude Opus 5Claude Sonnet 5Claude Fable 5Claude Mythos 5Claude Opus 4.8Claude Opus 4.7Claude Sonnet 4.6Claude Opus 4.6Claude Opus 4.5Claude Haiku 4.5Claude Sonnet 4.5Claude Opus 4.1Claude Opus 4Claude Sonnet 4Claude 3.7 SonnetClaude 3.5 HaikuClaude 3.5 SonnetClaude 3.5 SonnetClaude 3 HaikuClaude 3 OpusClaude 3 SonnetClaude Mythos Preview
Gemini 3.7 FlashGemini 3.5 Flash-LiteGemini 3.6 FlashGemini 3.5 FlashGemma 4 26B-A4BGemma 4 31BGemini 3.1 Flash-LiteGemini 3.1 ProGemini 3 FlashGemini 3 ProGemma 3n E4B InstructedGemini 2.5 Flash-LiteGemini 2.5 Pro Preview 06-05Gemini 2.5 FlashGemini 2.5 ProGemma 3 12BGemma 3 27BGemma 3 4BGemini 2.0 Flash-LiteGemini 2.0 FlashGemma 2 27BGemini 1.5 FlashGemini 1.5 ProGemini 1.5 Flash 8BGemini 1.0 Pro
Cursor
2 modelsxAI
15 modelsDeepSeek
16 modelsQwen
38 modelsQwen3.8-27BQwen3.8-2.4T-A95BQwen3.8 MaxQwen3.7 FlashQwen3.7-PlusQwen3.7 MaxQwen3.6-27BQwen3.6-35B-A3BQwen3.6 PlusQwen3.5-9BQwen3.5-122B-A10BQwen3.5-27BQwen3.5-35B-A3BQwen3.5-397B-A17BQwen3 MaxQwen3 VL 235B A22B InstructQwen3 VL 235B A22B ThinkingQwen3 VL 30B A3B InstructQwen3 VL 30B A3B ThinkingQwen3 VL 32B InstructQwen3 VL 4B InstructQwen3 VL 4B ThinkingQwen3 VL 8B InstructQwen3 VL 8B ThinkingQwen3-Next-80B-A3B-InstructQwen3-Next-80B-A3B-ThinkingQwen3-235B-A22B-Thinking-2507Qwen3-235B-A22B-Instruct-2507Qwen3 235B A22BQwen3 30B A3BQwen3 32BQwen3-Coder 480B A35B InstructQwen2.5 VL 72B InstructQwen3-CoderQwQ-32B-PreviewQwen2.5 72B InstructQwen2.5 7B InstructQwen2.5-Coder 32B Instruct
AI21 Labs
2 models
Amazon
5 models
Baidu
1 modelsByteDance
2 models
Cohere
3 models
IBM
1 models
Inception
1 models
inclusionAI
1 models
LG AI Research
1 models
Meituan
4 modelsMeta
12 modelsMistral
23 modelsMistral Medium 3.5Mistral Small 4Min istral 3 (3B Reasoning 2512)MiniStral 3 (14B Instruct 2512)Ministral 3 (14B Reasoning 2512)Ministral 3 (3B Instruct 2512)Ministral 3 (8B Instruct 2512)Ministral 3 (8B Reasoning 2512)Mistral Large 3 (675B Instruct 2512)Mistral Large 3Devstral Small 1.1Devstral MediumMistral Small 3.2 24B InstructMagistral MediumMistral Small 3.1 24B BaseMistral Small 3.1 24B InstructMistral Small 3 24B InstructPixtral LargeMinistral 8B InstructMistral SmallPixtral-12BMistral Large 2Mistral NeMo Instruct
Moonshot AI
9 models
Nous Research
1 models
NVIDIA
5 models
Poolside
2 models
Sakana AI
1 models
StepFun
1 modelsTencent
1 modelsThinking Machines
1 models
Unisound
1 models
Upstage
1 modelsXiaomi
3 modelsWhat is the best AI for coding?
As of August 22, 2026, DeepSeek-V4-Pro-0813 by DeepSeek is #1 for coding at 96.4%. Ranked by the SWE-bench Verified score This board also tracks SWE-bench Verified, LiveCodeBench, SciCode, Arena Code Elo. Next on the same board: GPT-5.6 Sol and Claude Opus 5. Related leaders: DeepSeek-V4-Pro-Max on LiveCodeBench at 93.5%; Claude Fable 5 on SciCode at 60.2%. This coding leaderboard ranks models by SWE-bench Verified. Scores come from public evals. Prices are the live API rates in the table above.
Sources: SWE-bench Verified (https://www.swebench.com/verified.html); OpenAI API pricing (https://developers.openai.com/api/docs/pricing); Anthropic Claude API pricing (https://platform.claude.com/docs/en/about-claude/pricing); Gemini API pricing (https://ai.google.dev/gemini-api/docs/pricing)
| Rank | Model | SWE-bench Verified | Input /M | Output /M |
|---|---|---|---|---|
| 1 | DeepSeek-V4-Pro-0813 | 96.4% | $1.32 | $3.96 |
| 2 | GPT-5.6 Sol | 96.2% | $5.00 | $30.00 |
| 3 | Claude Opus 5 | 96% | $5.00 | $25.00 |
| 4 | Grok 4.6 | 95.6% | $2.00 | $6.00 |
| 5 | Claude Mythos 5 | 95.5% | $10.00 | $50.00 |
| 6 | Claude Fable 5 | 95% | $10.00 | $50.00 |
| 7 | Claude Mythos Preview | 93.9% | $10.00 | $50.00 |
| 8 | Kimi K3 | 93.4% | $3.00 | $15.00 |
What is SWE-bench Verified?
SWE-bench Verified is 500 real GitHub issues, human-filtered so the tests are fair. Labs quote it because HumanEval saturated. A high Verified score means the model can patch a repo, not that it writes pretty snippets. Check Terminal-Bench if your agent lives in the shell.
SWE-bench vs Terminal-Bench
SWE-bench is repository issue fixing. Terminal-Bench is terminal and shell work. Claude often looks stronger on repo patches. GPT-class models often look stronger in the terminal. If your product is an agent, sort both columns before you lock a default model.
Claude vs ChatGPT for coding
Do not pick a brand. Rank the specific Claude and GPT rows on SWE-bench Verified and Terminal-Bench, then look at $/M. Cursor vs Claude is the wrong comparison: Cursor is the editor, Claude is one of the models it can call.
Related LLM leaderboards
Individual benchmark pages
Rank one eval at a time. All LLM benchmarks.
Coding FAQ
What is the best AI for coding?
As of August 22, 2026, DeepSeek-V4-Pro-0813 by DeepSeek is #1 for coding at 96.4%. Ranked by the SWE-bench Verified score This board also tracks SWE-bench Verified, LiveCodeBench, SciCode, Arena Code Elo. Next on the same board: GPT-5.6 Sol and Claude Opus 5. Related leaders: DeepSeek-V4-Pro-Max on LiveCodeBench at 93.5%; Claude Fable 5 on SciCode at 60.2%.
What are the top models on SWE-bench Verified?
The current SWE-bench Verified ranking as of August 22, 2026 is 1. DeepSeek-V4-Pro-0813 at 96.4%; 2. GPT-5.6 Sol at 96.2%; 3. Claude Opus 5 at 96%.
Which coding model is the cheapest?
Qwen2.5-Coder 32B Instruct is the cheapest scored model on this coding leaderboard at $0.09/M input and $0.09/M output ($0.18 blended). DeepSeek-V4-Pro-0813 still leads SWE-bench Verified at 96.4%.
Should I always pick the #1 SWE-bench Verified model?
Not automatically. DeepSeek-V4-Pro-0813 leads SWE-bench Verified, but a cheaper scored model can be the better production choice if the quality gap is small. Use the table to weigh SWE-bench Verified against input/output price, context window, and related evals.
How often is the Coding leaderboard updated?
Scores and API prices on this page are refreshed from published evals and provider rates. The snapshot is labeled August 22, 2026. Treat it as a current index, not a one-off blog post.
What is the best LLM for coding right now?
As of August 22, 2026, DeepSeek-V4-Pro-0813 by DeepSeek leads this coding leaderboard on SWE-bench Verified at 96.4%. API pricing is $1.32/M input and $3.96/M output. If a cheaper scored model is close on SWE-bench, that is often the better production pick.
What is the best AI for coding right now?
The best coding model depends on the eval. SWE-bench Verified measures real GitHub fixes, LiveCodeBench tracks contest programming, and Terminal-Bench tests shell agents. Rank the table by the job you actually run.
Is SWE-bench the same as LiveCodeBench?
No. SWE-bench scores repository-level issue fixing. LiveCodeBench scores fresh competitive-programming problems. A model can lead one and lag the other.
Do coding leaderboard scores include API price?
Yes. Every row shows input and output token prices so you can see whether the top SWE-bench model is worth the extra cost versus an open-source coder.
What is the best coding AI?
Same question as best AI for coding. Rank SWE-bench Verified for GitHub issue fixes and Terminal-Bench for shell agents. A model can lead one and lag the other. Price sits on the same row so you can skip a tiny score gap that triples the bill.
What is SWE-bench?
SWE-bench scores whether a model can resolve real GitHub issues. SWE-bench Verified is the human-filtered 500-issue subset most labs quote. SWE-bench Pro and DeepSWE are harder follow-ons. HumanEval is a short function-completion test, not the same job.
What is SWE-bench Verified?
SWE-bench Verified is 500 human-checked GitHub issues. It is the default public coding rank in 2026. Bash-only and full-agent harnesses are not interchangeable, so compare scores that name the same setup.
What is Terminal-Bench?
Terminal-Bench scores shell and terminal agents: install tools, run commands, finish a task in a real environment. That is closer to a coding agent loop than HumanEval. SWE-bench is repo patches. Terminal-Bench is the terminal. Sort both.
Claude vs ChatGPT for coding: which is better?
Claude vs ChatGPT for coding is a SKU fight. Sonnet/Opus-class Claude and GPT-class OpenAI models trade SWE-bench and Terminal-Bench leads as evals refresh. Rank this table, then take the cheaper model if the gap is small.
Cursor vs Claude: which is better for coding?
Cursor is an IDE that calls models, often Claude. Claude is the model. Ranking Cursor vs Claude mixes a product and a model. Pick the model on this SWE-bench board, then pick the client (Cursor, Codex, Claude Code) separately.
What is the best AI coding assistant?
An AI coding assistant is the client plus the model. This page ranks the models on SWE-bench Verified, Terminal-Bench, LiveCodeBench, and related evals, with API price. The assistant UI is a different purchase.
What is the best open source LLM for coding?
Open the open-source board and sort by SWE-bench or LiveCodeBench. Qwen, DeepSeek, Kimi, and Llama variants rotate. Weights can be free. The hosted coding API still bills tokens.