# gpqa-diamond benchmark

| \# | model | gpqa-diamond | $/task | card | cheapest list | cheapest effective | unit | providers | tokens/week |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
| 1 | `google/gemini-4-argon` | 96.5 | 0.0947 |  | 1.0 google-ai-studio/flex | 0.4977 google-ai-studio/flex |  | 2 | 33357280151 |
| 2 | `google/gemini-3.8-flash` | 95.6 | 0.0682 |  | 0.375 google-vertex/global/flex | 0.1745 google-ai-studio/flex |  | 2 | 66105151093 |
| 3 | `openai/gpt-6-astra-pro` | 95.5 | 0.3326 | 10.0 | 5.0 openai/flex | 1.4761 openai/flex |  | 2 | 1273455495 |
| 4 | `google/gemini-3.1-pro-preview` | 94.5 | 0.202 |  | 1.0 google-ai-studio/flex | 0.6497 google-ai-studio/flex |  | 2 | 4366256457 |
| 5 | `openai/gpt-6.1-sol` | 94.4 | 0.0214 | 2.0 | 1.0 openai/flex | 0.111 openai/flex |  | 3 | 120159205745 |
| 6 | `openai/gpt-6-astra` | 94.4 | 0.1133 | 10.0 | 5.0 openai/flex | 1.0295 openai/flex |  | 3 | 12764530937 |
| 7 | `openai/gpt-5.6-sol-pro` | 94.1 | 0.1513 | 4.0 | 2.0 openai/flex | 1.4685 openai |  | 2 | 695188392 |
| 8 | `openai/gpt-5.5` | 93.9 | 0.3125 | 5.0 | 2.5 openai/flex | 0.9605 openai/flex |  | 3 | 2430328494 |
| 9 | `google/gemini-3.7-flash` | 93.9 | 0.0293 |  | 0.375 google-vertex/global/flex | 0.1294 google-ai-studio/flex |  | 2 | 16485676436 |
| 10 | `moonshotai/kimi-k3.1-pro` | 92.9 | 0.1116 | 2.0 | 0.304 relace/fp4 | 0.2079 relace/fp4 |  | 17 | 127045171781 |
| 11 | `x-ai/grok-4.6` | 92.8 | 0.1233 | 2.0 | 2.0 xai/zdr | 0.7725 amazon-bedrock/us-west-2 |  | 2 | 2750565342 |
| 12 | `anthropic/claude-sonnet-5.5` | 92.8 | 0.0252 | 2.0 | 2.0 amazon-bedrock | 0.3958 google-vertex/us |  | 5 | 90350909445 |
| 13 | `google/gemini-3.6-flash` | 92.8 | 0.0444 |  | 0.375 google-ai-studio/flex | 0.3217 google-ai-studio/flex |  | 2 | 3715265581 |
| 14 | `openai/gpt-5.6-sol` | 92.4 | 0.0386 | 4.0 | 2.0 openai/flex | 1.2602 azure/us |  | 3 | 10412992133 |
| 15 | `google/gemini-3.5-flash` | 92.3 | 0.1437 |  | 0.75 google-ai-studio/flex | 0.2479 google-ai-studio/flex |  | 2 | 1763021737 |
| 16 | `openai/gpt-6-sol` | 91.9 | 0.0253 | 2.0 | 1.0 openai/flex | 0.4568 openai/flex |  | 3 | 8857381915 |
| 17 | `x-ai/grok-4.7` | 91.9 | 0.2183 | 2.0 | 2.0 amazon-bedrock/global | 0.6736 xai/zdr |  | 2 | 21521368346 |
| 18 | `moonshotai/kimi-k3` | 91.9 | 0.1286 | 3.0 | 0.153 wafer/fast | 0.1521 wafer/fast |  | 22 | 59504210550 |
| 19 | `openai/gpt-6.1-terra` | 91.1 | 0.0115 | 1.0 | 0.5 openai/flex | 0.0589 openai/flex |  | 3 | 44362382979 |
| 20 | `stealth/union-alpha` | 90.9 |  |  |  |  |  | 1 | 307542656848 |
| 21 | `qwen/qwen3.7-max` | 90.9 | 0.2222 |  | 1.475 alibaba | 0.6225 alibaba |  | 1 | 229410698 |
| 22 | `deepseek/deepseek-v4-pro-0813` | 90.3 | 0.106 | 1.32 | 0.0999 relace/fp4 | 0.0998 relace/fp4 |  | 24 | 38297722212 |
| 23 | `minimax/minimax-m3` | 90.3 | 0.0347 | 0.3 | 0.224 deepinfra/fp8/flex | 0.0623 gmicloud/fp8 |  | 14 | 42355642611 |
| 24 | `openai/gpt-5.4` | 90.2 | 0.1438 | 2.5 | 1.25 openai/flex | 0.9545 openai |  | 3 | 1559911228 |
| 25 | `anthropic/claude-opus-5.5` | 90.2 | 0.0541 | 4.0 | 4.0 azure/global | 0.3991 google-vertex/europe |  | 5 | 265225848809 |
| 26 | `meta/muse-spark-open` | 90.2 | 0.0737 |  | 1.25 baseten/fp4 | 0.2765 makora/fp4 |  | 5 | 3455713072 |
| 27 | `deepseek/deepseek-v4.1-flash-1017` | 90.1 | 0.011 | 0.3 | 0.006 relace/fp4 | 0.0045 deepinfra/fp8/flex |  | 41 | 2701423607321 |
| 28 | `anthropic/claude-fable-5.1` | 89.9 | 0.1305 | 10.0 | 10.0 amazon-bedrock | 2.1435 anthropic |  | 4 | 7573541962 |
| 29 | `tencent/hy4-preview` | 89.8 | 0.1192 | 0.834 | 0.834 atlas-cloud/fp8 | 0.0646 tencent/fp8 |  | 6 | 6944527610 |
| 30 | `anthropic/claude-opus-4.8` | 89.6 | 0.2002 | 5.0 | 5.0 azure/global | 1.1933 amazon-bedrock |  | 5 | 7736756673 |
| 31 | `openai/gpt-5.6-luna-pro` | 89.6 | 0.0384 | 0.2 | 0.1 openai/flex | 0.0654 openai/flex |  | 2 | 3137471618 |
| 32 | `anthropic/claude-opus-4.7` | 89.5 | 0.2322 | 5.0 | 5.0 google-vertex/global | 1.5151 anthropic |  | 5 | 874219417 |
| 33 | `deepseek/deepseek-v4.1-flash` | 89.1 | 0.023 | 0.15 | 0.006 relace/fp4 | 0.0048 primeintellect |  | 32 | 831323377790 |
| 34 | `xiaomi/mimo-v2.6-pro` | 88.8 | 0.0329 | 0.435 | 0.3262 gmicloud/bf16 | 0.0284 gmicloud/bf16 |  | 5 | 33262750560 |
| 35 | `amazon/nova-micro-v1` | 88.6 | 0.2216 |  | 0.035 amazon-bedrock/eu-west-1 | 0.0347 amazon-bedrock/eu-west-1 |  | 1 | 753428554 |
| 36 | `qwen/qwen3.8-flash` | 88.6 | 0.0215 |  | 0.15 alibaba | 0.0322 alibaba |  | 2 | 45074568765 |
| 37 | `google/gemini-3-flash-preview` | 88.5 | 0.132 |  | 0.25 google-ai-studio/flex | 0.0928 google-ai-studio/flex |  | 2 | 22369124012 |
| 38 | `reflection/beam-1.0-501b-a23b` | 88.4 | 0.0147 |  | 0.25 nebius/fp4 | 0.0612 together |  | 5 | 2111325689 |
| 39 | `openai/gpt-6-luna-pro` | 88.4 | 0.0083 | 0.1 | 0.05 openai/flex | 0.0201 openai/flex |  | 2 | 20728381439 |
| 40 | `openai/gpt-5.6-terra` | 88.4 | 0.0413 | 2.0 | 1.0 openai/flex | 0.4604 azure |  | 3 | 4625799877 |
| 41 | `deepseek/deepseek-v4-flash-vision-exp` | 88.0 | 0.0228 | 0.44 | 0.2156 deepinfra/fp8 | 0.0254 gmicloud/fp8 |  | 7 | 19162881398 |
| 42 | `anthropic/claude-opus-5` | 87.9 | 0.128 | 5.0 | 5.0 azure/global | 1.1673 anthropic |  | 5 | 9999518143 |
| 43 | `qwen/qwen3.7-plus` | 87.9 | 0.044 |  | 0.32 alibaba | 0.1732 alibaba |  | 1 | 2491495658 |
| 44 | `openai/gpt-5.2` | 87.9 | 0.1333 | 1.75 | 0.875 openai/flex | 0.6736 openai/flex |  | 2 | 1003013452 |
| 45 | `minimax/minimax-m3.1-flash` | 87.6 | 0.014 | 0.15 | 0.1125 gmicloud/fp8 | 0.0186 gmicloud/fp8 |  | 12 | 165405568604 |
| 46 | `openai/gpt-5.6-luna` | 87.6 | 0.0094 | 0.2 | 0.1 openai/flex | 0.0591 openai/flex |  | 3 | 65706471235 |
| 47 | `qwen/qwen3.8-2.4t-a95b` | 87.4 | 0.0905 |  | 2.0 alibaba | 0.3546 together |  | 9 | 3375846448 |
| 48 | `openai/gpt-6-luna` | 87.4 | 0.003 | 0.1 | 0.05 openai/flex | 0.0307 azure |  | 3 | 307311562986 |
| 49 | `anthropic/claude-fable-5` | 87.3 | 0.1821 | 10.0 | 10.0 azure | 3.27 anthropic |  | 5 | 1799948098 |
| 50 | `deepseek/deepseek-v4-pro` | 87.1 | 0.0447 | 0.66 | 0.66 deepseek | 0.1881 streamlake/fp8 |  | 21 | 58724115254 |
| 51 | `nvidia/nemotron-3.5-super-vl` | 86.8 | 0.0219 |  | 0.06 darkbloom/nvfp4 | 0.0344 primeintellect/nvfp4 |  | 10 | 18683979833 |
| 52 | `anthropic/claude-haiku-5.5` | 86.7 | 0.0041 | 0.1 | 0.1 anthropic | 0.0425 azure/global |  | 5 | 139788290103 |
| 53 | `openai/gpt-5.1` | 86.7 | 0.2227 | 1.25 | 0.625 openai/flex | 0.2471 openai/flex |  | 2 | 999470698 |
| 54 | `moonshotai/kimi-k2.7-code` | 86.7 | 0.1023 | 0.95 | 0.6 coreweave/int4 | 0.2388 coreweave/int4 |  | 18 | 1525673246 |
| 55 | `deepseek/deepseek-v4-flash-0731` | 86.6 | 0.0077 | 0.44 | 0.03 open-inference/fp4 | 0.0139 streamlake/fp8 |  | 37 | 314513489706 |
| 56 | `thinkingmachines/inkling-small` | 86.6 | 0.0253 |  | 0.45 deepinfra/fp8 | 0.1244 baseten/fp8 |  | 3 | 646710202 |
| 57 | `anthropic/claude-opus-4.6` | 86.4 | 0.75 | 5.0 | 5.0 google-vertex/global | 1.5218 anthropic |  | 5 | 4491236099 |
| 58 | `anthropic/claude-opus-4.5` | 86.4 | 0.8426 | 5.0 | 5.0 anthropic/2 | 3.6279 anthropic |  | 5 | 219516097 |
| 59 | `moonshotai/kimi-k2.6` | 86.3 | 0.1642 | 0.95 | 0.39 chutes/int4 | 0.2277 chutes/int4 |  | 22 | 4176548547 |
| 60 | `qwen/qwen3.5-397b-a17b` | 85.8 | 0.0787 |  | 0.39 alibaba | 0.3199 deepinfra/fp8 |  | 11 | 511195664 |
| 61 | `openai/gpt-5` | 85.7 | 0.2099 | 1.25 | 1.25 azure | 0.7591 openai/default |  | 2 | 311907360 |
| 62 | `deepseek/deepseek-v4-flash` | 85.6 | 0.0058 | 0.22 | 0.05 relace/fp4 | 0.0203 open-inference/fp8 |  | 24 | 264245760211 |
| 63 | `minimax/minimax-m2.7` | 85.6 | 0.0461 | 0.3 | 0.21 gmicloud/fp8 | 0.0903 deepinfra/fp8 |  | 9 | 4062222361 |
| 64 | `z-ai/glm-5.3` | 85.4 | 0.1033 | 1.4 | 0.0699 wafer/us | 0.069 wafer/us |  | 39 | 244459012685 |
| 65 | `anthropic/claude-sonnet-5` | 85.2 | 0.1692 | 2.0 | 2.0 amazon-bedrock/global | 0.3115 google-vertex/us |  | 5 | 19785018514 |
| 66 | `minimax/minimax-m2.1` | 85.2 | 0.0168 | 0.3 | 0.3 minimax/fp8 | 0.0952 minimax/fp8 |  | 2 | 300923844 |
| 67 | `z-ai/glm-5.2` | 84.9 | 0.1367 | 1.4 | 0.0066 relace/fp4 | 0.0062 wafer/us |  | 34 | 67682870933 |
| 68 | `inclusionai/ling-3.0-flash-vl` | 84.8 | 0.0013 |  | 0.06 deepinfra/fp16 | 0.028 deepinfra/fp16 |  | 1 | 144990235 |
| 69 | `moonshotai/kimi-k2.5` | 84.8 | 0.0988 | 0.6 | 0.45 deepinfra/fp4 | 0.1505 moonshotai/int4 |  | 10 | 1480447887 |
| 70 | `google/gemini-3.5-flash-lite` | 84.4 | 0.0364 |  | 0.15 google-ai-studio/flex | 0.093 google-vertex/global/flex |  | 2 | 8815470911 |
| 71 | `qwen/qwen3.5-122b-a10b` | 84.3 | 0.0544 |  | 0.26 alibaba | 0.2597 siliconflow/fp8 |  | 5 | 456557296 |
| 72 | `z-ai/glm-5.3-flash` | 84.1 | 0.0135 | 0.15 | 0.0399 relace/fp4 | 0.0204 primeintellect |  | 40 | 806116523144 |
| 73 | `openai/gpt-5.4-mini` | 83.8 | 0.0619 | 0.75 | 0.375 openai/flex | 0.0599 openai/flex |  | 2 | 3592057033 |
| 74 | `moonshotai/kimi-k2-thinking` | 83.4 | 0.0461 |  | 0.6 google-vertex | 0.3613 novita/bf16 |  | 2 | 101424456 |
| 75 | `thinkingmachines/inkling` | 83.4 | 0.0954 |  | 0.95 deepinfra/fp8 | 0.6663 together |  | 3 | 4073704634 |
| 76 | `qwen/qwen3.6-27b` | 82.8 | 0.0889 |  | 0.25 chutes/fp8 | 0.1004 chutes/fp8 |  | 10 | 1018016186 |
| 77 | `google/gemini-2.5-pro` | 82.6 | 0.2609 |  | 0.625 google-ai-studio/flex | 0.2859 google-ai-studio/flex |  | 2 | 4353720545 |
| 78 | `xiaomi/mimo-v2.5-pro` | 82.4 | 0.0287 | 0.435 | 0.3045 gmicloud/bf16 | 0.0612 xiaomi/fp8 |  | 7 | 17728731871 |
| 79 | `nvidia/nemotron-3-ultra-550b-a55b` | 82.2 | 0.083 |  | 0.4 deepinfra/fp4/flex | 0.1215 deepinfra/fp4/flex |  | 5 | 3907102956 |
| 80 | `google/gemini-3.1-flash-lite` | 82.1 | 0.0406 |  | 0.125 google-vertex/global/flex | 0.0667 google-vertex/eu |  | 2 | 17408115717 |
| 81 | `google/gemma-4-31b-it` | 82.0 | 0.0075 |  | 0.08 open-inference/bf16 | 0.0798 open-inference/bf16 |  | 22 | 24673261769 |
| 82 | `qwen/qwen3.8-27b` | 81.9 | 0.0336 |  | 0.035 darkbloom/fp4 | 0.0103 chutes/fp8 |  | 26 | 38376372463 |
| 83 | `z-ai/glm-4.7` | 81.8 | 0.0892 | 0.6 | 0.32 deepinfra/fp4/flex | 0.2166 deepinfra/fp4/flex |  | 9 | 1793109277 |
| 84 | `anthropic/claude-sonnet-4.5` | 81.7 | 0.3136 | 3.0 | 3.0 google-vertex/global | 0.6022 google-vertex/us-east5 |  | 5 | 2446063053 |
| 85 | `inclusionai/ling-3.1-flash` | 81.3 | 0.0077 |  | 0.045 deepinfra/fp4 | 0.0186 novita |  | 2 | 2465499085 |
| 86 | `deepseek/deepseek-v3.2` | 81.0 | 0.0093 |  | 0.2088 gmicloud/fp8 | 0.0926 gmicloud/fp8 |  | 15 | 29736505094 |
| 87 | `openai/gpt-5-mini` | 81.0 | 0.045 | 0.25 | 0.125 openai/flex | 0.0405 openai/flex |  | 2 | 2841429611 |
| 88 | `z-ai/glm-5.1` | 80.8 | 0.2363 | 1.4 | 0.9 chutes/fp8 | 0.3379 chutes/fp8 |  | 18 | 1682206009 |
| 89 | `meta/muse-glimmer-30b` | 80.6 | 0.0225 |  | 0.3 deepinfra/bf16 | 0.0604 phala |  | 5 | 1024965298 |
| 90 | `minimax/minimax-m2.5` | 80.4 | 0.0288 | 0.3 | 0.22 inceptron/fp8 | 0.0452 siliconflow/fp8 |  | 11 | 2024556517 |
| 91 | `qwen/qwen3.5-35b-a3b` | 80.1 | 0.0267 |  | 0.065 darkbloom/fp4 | 0.065 darkbloom/fp4 |  | 9 | 1631869476 |
| 92 | `openai/gpt-5.2-chat` | 79.8 | 0.036 |  | 1.75 azure |  |  | 1 |  |
| 93 | `qwen/qwen3.6-35b-a3b` | 79.7 | 0.0431 |  | 0.018 primeintellect/fp8 | 0.0127 primeintellect/fp8 |  | 12 | 9059542464 |
| 94 | `deepseek/deepseek-v3.2-exp` | 78.9 | 0.0044 |  | 0.27 atlas-cloud/fp8 | 0.2699 atlas-cloud/fp8 |  | 3 | 1057454921 |
| 95 | `anthropic/claude-sonnet-4` | 78.6 | 0.2643 |  | 3.0 amazon-bedrock | 1.4411 google-vertex/global |  | 2 | 1266922610 |
| 96 | `xiaomi/mimo-v2.5` | 77.8 | 0.0133 | 0.14 | 0.119 gmicloud/fp8 | 0.0126 xiaomi/fp8 |  | 7 | 490191247748 |
| 97 | `openai/gpt-5.4-nano` | 77.7 | 0.0131 | 0.2 | 0.1 openai/flex | 0.0491 openai/flex |  | 2 | 2347263828 |
| 98 | `qwen/qwen3.5-9b` | 77.5 | 0.0059 |  | 0.053 primeintellect/bf16 | 0.0787 darkbloom/fp4 |  | 7 | 3637842369 |
| 99 | `z-ai/glm-5` | 77.4 | 0.1597 | 1.0 | 0.6 streamlake/fp8 | 0.2836 z-ai/fp8 |  | 11 | 1644894669 |
| 100 | `stepfun/step-3.7-flash` | 76.6 | 0.0802 | 0.2 | 0.16 deepinfra | 0.0456 deepinfra |  | 3 | 750713750 |
| 101 | `deepseek/deepseek-r1-0528` | 76.5 | 0.052 |  | 0.5 deepinfra/fp4 | 0.4381 deepinfra/fp4 |  | 4 | 426276956 |
| 102 | `deepseek/deepseek-v3.1-terminus` | 76.3 | 0.0172 |  | 0.27 deepinfra/fp4 | 0.2037 atlas-cloud/fp8 |  | 5 | 823280489 |
| 103 | `qwen/qwen3-235b-a22b-thinking-2507` | 76.2 | 0.0547 |  | 0.23 alibaba | 0.2299 alibaba |  | 4 | 251351864 |
| 104 | `z-ai/glm-4.6` | 76.1 | 0.0587 | 0.6 | 0.4 deepinfra/fp4/flex | 0.163 venice/fp4 |  | 5 | 854974071 |
| 105 | `moonshotai/kimi-k2-0905` | 75.9 | 0.0098 |  | 0.6 novita/fp8 | 0.5999 novita/fp8 |  | 1 | 225119163 |
| 106 | `mistralai/mistral-small-2603` | 75.8 | 0.0146 | 0.15 | 0.15 mistral | 0.0662 mistral/zdr |  | 2 | 1133534667 |
| 107 | `qwen/qwen3-coder-next` | 74.9 | 0.0148 |  | 0.12 parasail/bf16 | 0.0733 parasail/bf16 |  | 4 | 1276802354 |
| 108 | `inclusionai/ling-3.0-flash` | 74.9 | 0.0026 |  | 0.021 novita | 0.0123 novita |  | 2 | 4782103295 |
| 109 | `openai/gpt-oss-120b` | 73.6 | 0.0067 |  | 0.03 akashml/fp4 | 0.0204 near-ai/fp4 |  | 24 | 16824607232 |
| 110 | `xiaomi/mimo-v2.6-flash` | 73.2 | 0.0165 | 0.14 | 0.1 io-net/fp8 | 0.011 gmicloud/bf16 |  | 9 | 188589602106 |
| 111 | `anthropic/claude-haiku-4.5` | 73.0 | 0.1904 | 1.0 | 1.0 azure/global | 0.2325 amazon-bedrock/eu-west-1 |  | 4 | 5580791996 |
| 112 | `google/gemini-2.5-flash` | 72.8 | 0.0608 |  | 0.15 google-ai-studio/flex | 0.0958 google-ai-studio/flex |  | 2 | 39453554708 |
| 113 | `qwen/qwen3-235b-a22b-2507` | 72.6 | 0.0055 |  | 0.0875 gmicloud/fp8 | 0.0897 deepinfra/fp8 |  | 11 | 3006307426 |
| 114 | `qwen/qwen3-next-80b-a3b-instruct` | 71.8 | 0.011 |  | 0.072 deepinfra/fp8/flex | 0.0719 deepinfra/fp8/flex |  | 5 | 188711750 |
| 115 | `anthropic/claude-sonnet-4.6` | 70.6 | 0.6159 | 3.0 | 3.0 google-vertex/global | 1.072 amazon-bedrock/global |  | 5 | 9270256352 |
| 116 | `openai/gpt-5-nano` | 70.5 | 0.0134 | 0.05 | 0.025 openai/flex | 0.0169 openai/flex |  | 2 | 2151750077 |
| 117 | `nvidia/nemotron-3.5-lightning` | 70.4 | 0.0091 |  | 0.025 darkbloom/int4 | 0.0175 darkbloom/int4 |  | 8 | 947152601 |
| 118 | `google/gemma-4-26b-a4b-it` | 70.2 | 0.015 |  | 0.01 primeintellect/bf16 | 0.0104 primeintellect/bf16 |  | 15 | 17632526472 |
| 119 | `qwen/qwen3-vl-235b-a22b-instruct` | 68.4 | 0.0096 |  | 0.2 deepinfra/fp8 | 0.145 parasail/fp8 |  | 5 | 1664163900 |
| 120 | `z-ai/glm-4.6v` | 66.8 | 0.0139 | 0.3 | 0.3 novita/bf16 | 0.2266 z-ai/fp8 |  | 2 | 79610746 |
| 121 | `meta-llama/llama-4-maverick` | 66.7 | 0.0029 |  | 0.1875 digitalocean | 0.1872 digitalocean |  | 5 | 1166480972 |
| 122 | `ibm-granite/granite-4.2-8b` | 65.7 | 0.0149 |  | 0.06 coreweave/bf16 | 0.0212 coreweave/bf16 |  | 2 | 92131414 |
| 123 | `deepseek/deepseek-chat-v3.1` | 65.4 | 0.0148 |  | 0.25 deepinfra/fp4 | 0.1968 deepinfra/fp4 |  | 8 | 3439868728 |
| 124 | `openai/gpt-4.1-mini` | 65.2 | 0.0042 | 0.4 | 0.4 azure | 0.2386 azure |  | 2 | 4293262180 |
| 125 | `openai/gpt-oss-20b` | 65.1 | 0.0059 |  | 0.018 darkbloom/fp8 | 0.0152 darkbloom/fp8 |  | 15 | 5890505215 |
| 126 | `qwen/qwen3-30b-a3b-instruct-2507` | 64.8 | 0.0028 |  | 0.0481 streamlake | 0.0479 streamlake |  | 6 | 2437832339 |
| 127 | `z-ai/glm-4.5-air` | 64.7 | 0.0126 | 0.2 | 0.13 novita/bf16 | 0.0762 novita/bf16 |  | 3 | 106946576 |
| 128 | `qwen/qwen3-vl-30b-a3b-instruct` | 64.0 | 0.0072 |  | 0.13 alibaba | 0.1291 alibaba |  | 4 | 445512644 |
| 129 | `openai/gpt-4.1` | 63.7 | 0.0155 | 2.0 | 2.0 azure | 1.2994 azure/swedencentral |  | 2 | 1682583922 |
| 130 | `nvidia/nemotron-3-nano-30b-a3b` | 62.9 | 0.0179 |  | 0.05 crusoe/fp8 | 0.0295 nebius/fp8 |  | 4 | 795814817 |
| 131 | `deepseek/deepseek-chat` | 62.1 | 0.0024 |  | 0.256 deepinfra/fp4/flex | 0.2538 deepinfra/fp4/flex |  | 3 | 795313498 |
| 132 | `qwen/qwen3-30b-a3b` | 60.9 | 0.0078 |  | 0.12 deepinfra/fp8 | 0.1198 deepinfra/fp8 |  | 2 | 220782763 |
| 133 | `qwen/qwen3-coder` | 60.7 | 0.0035 |  | 0.22 google-vertex/us-south1 | 0.0767 venice/fp8 |  | 5 | 595791409 |
| 134 | `qwen/qwen3-32b` | 60.4 | 0.0083 |  | 0.08 deepinfra/fp8 | 0.0795 deepinfra/fp8 |  | 5 | 1151022867 |
| 135 | `qwen/qwen3-14b` | 59.2 | 0.0029 |  | 0.1 nextbit/int4 | 0.0994 nextbit/int4 |  | 3 | 100754729 |
| 136 | `deepseek/deepseek-chat-v3-0324` | 56.3 | 0.0054 |  | 0.24 deepinfra/fp4 | 0.2182 deepinfra/fp4 |  | 5 | 1547656279 |
| 137 | `google/gemini-2.5-flash-lite` | 55.2 | 0.0178 |  | 0.05 google-ai-studio/flex | 0.0375 google-ai-studio/flex |  | 2 | 55259395661 |
| 138 | `qwen/qwen3-vl-8b-instruct` | 53.9 | 0.0055 |  | 0.117 alibaba | 0.1167 alibaba |  | 2 | 511387540 |
| 139 | `z-ai/glm-4.7-flash` | 52.2 | 0.0169 |  | 0.06 deepinfra/bf16 | 0.0279 deepinfra/bf16 |  | 4 | 503434086 |
| 140 | `qwen/qwen3-coder-30b-a3b-instruct` | 52.2 | 0.0043 |  | 0.07 novita/fp8 | 0.0699 siliconflow/fp8 |  | 5 | 267638432 |
| 141 | `openai/gpt-4o-2024-08-06` | 52.0 | 0.0175 | 2.5 | 2.5 azure | 2.2318 azure |  | 2 | 48293294 |
| 142 | `openai/gpt-4o` | 51.1 | 0.0132 | 2.5 | 2.5 azure | 2.2567 openai |  | 2 | 531370864 |
| 143 | `openai/gpt-4o-2024-05-13` | 50.7 | 0.0237 | 5.0 | 5.0 azure | 5.0 azure |  | 2 | 4475376 |
| 144 | `inclusionai/ling-3.0-flash-fin` | 50.5 | 0.0098 |  | 0.06 deepinfra/fp4 | 0.0169 deepinfra/fp4 |  | 1 | 264897851 |
| 145 | `openai/gpt-4.1-nano` | 49.1 | 0.0008 | 0.1 | 0.1 azure | 0.0628 azure/swedencentral |  | 2 | 5749479513 |
| 146 | `meta-llama/llama-3.3-70b-instruct` | 47.6 | 0.0014 |  | 0.1 deepinfra/turbo | 0.0997 deepinfra/turbo |  | 12 | 1923672456 |
| 147 | `qwen/qwen-2.5-72b-instruct` | 46.0 | 0.001 |  | 0.36 deepinfra/fp8 | 0.3595 deepinfra/fp8 |  | 2 | 42330708 |
| 148 | `openai/gpt-4o-mini` | 43.6 | 0.001 | 0.15 | 0.15 azure | 0.1112 azure |  | 2 | 7411041954 |
| 149 | `qwen/qwen2.5-vl-72b-instruct` | 42.9 | 0.0016 |  | 0.25 nebius/fp8 | 0.2499 nebius/fp8 |  | 2 | 141573502 |
| 150 | `tencent/hy-mt2-30b-a3b` | 38.9 | 0.0002 | 0.074 | 0.074 tencent/fp8 | 0.0723 tencent/fp8 |  | 1 | 51188164 |
| 151 | `google/gemma-3-12b-it` | 38.0 | 0.0004 |  | 0.04 deepinfra/bf16/flex | 0.0394 deepinfra/bf16/flex |  | 2 | 348865025 |
| 152 | `qwen/qwen-2.5-7b-instruct` | 32.8 | 0.0005 |  | 0.1 phala | 0.0996 phala |  | 2 | 343255240 |
| 153 | `mistralai/mistral-nemo` | 31.6 | 0.0001 | 0.15 | 0.0166 io-net/fp16 | 0.0122 io-net/fp16 |  | 6 | 15261188711 |
| 154 | `meta-llama/llama-3.2-3b-instruct` | 28.6 | 0.0006 |  | 0.05 parasail/bf16 | 0.0478 parasail/bf16 |  | 2 | 54620412 |
| 155 | `meta-llama/llama-3.1-8b-instruct` | 27.9 | 0.0011 |  | 0.02 deepinfra/fp8 | 0.0194 novita/fp8 |  | 5 | 1886655330 |
| 156 | `sao10k/l3-lunaris-8b` | 27.7 | 0.0001 |  | 0.04 deepinfra/turbo | 0.0398 parasail/bf16 |  | 3 | 243678333 |
| 157 | `google/gemma-3-4b-it` | 27.1 | 0.0001 |  | 0.04 parasail/bf16 | 0.0366 parasail/bf16 |  | 2 | 227868747 |
| 158 | `google/gemini-3.1-flash-lite-image` | 10.3 | 0.018 |  | 0.25 google-ai-studio | 0.2492 google-vertex/global |  | 2 | 55315829 |