LLM Optimizer

Find the smartest model for your budget

Data updated 424h 37m ago from Artificial Analysis · 102 models

Model Creator Intelligence Coding Input $/M Output $/M Blended $/M Speed t/s TTFT s Context Value ↓
Claude Haiku 5.5 (Max) Anthropic 36.5 — $0.10 $0.50 $0.20 236.3 256.31 — 182.5
Claude Haiku 5.5 (Xhigh) Anthropic 34.6 — $0.10 $0.50 $0.20 174.0 50.47 — 173.0
MiMo-V2.6-Flash Xiaomi 26.7 — $0.14 $0.28 $0.17 29.7 2.43 — 152.6
Claude Haiku 5.5 (High) Anthropic 29.9 — $0.10 $0.50 $0.20 148.6 16.25 — 149.5
GPT-6 Luna (Max) OpenAI 27.9 — $0.10 $0.50 $0.20 123.3 110.74 — 139.5
Claude Haiku 5.5 (Medium) Anthropic 24.6 — $0.10 $0.50 $0.20 143.8 9.00 — 123.0
GPT-6 Luna (Xhigh) OpenAI 23.5 — $0.10 $0.50 $0.20 115.9 21.08 — 117.5
GLM-5.3-Flash Z AI 27.8 — $0.15 $0.50 $0.24 64.1 1.04 — 117.3
MiniMax-M3.1-Flash MiniMax 29.1 — $0.15 $0.60 $0.26 107.9 0.74 — 111.1
GPT-6 Luna (High) OpenAI 21.4 — $0.10 $0.50 $0.20 117.8 9.96 — 107.0
Claude Haiku 5.5 (Low) Anthropic 18.6 — $0.10 $0.50 $0.20 142.2 5.86 — 93.0
GPT-6 Luna (Medium) OpenAI 18.5 — $0.10 $0.50 $0.20 0.0 0.00 — 92.5
Ling 3.1 Flash 1019 InclusionAI 29.6 — $0.30 $0.90 $0.45 0.0 0.00 — 65.8
DeepSeek V4.1 Flash 1017 (Max) DeepSeek 32.6 — $0.30 $1.20 $0.53 155.4 0.88 — 62.1
Nemotron 3.5 Lightning NVIDIA 5.5 — $0.06 $0.20 $0.10 270.4 0.37 — 57.9
MiMo-V2.6-Pro Xiaomi 29.9 — $0.43 $0.87 $0.54 38.4 2.34 — 55.0
DeepSeek V4.1 Flash 1017 (High) DeepSeek 28.6 — $0.30 $1.20 $0.53 165.9 0.84 — 54.5
GPT-5.6 Luna (Max) OpenAI 23.8 — $0.20 $1.20 $0.45 0.0 0.00 — 52.9
Nemotron 3.5 Super VL NVIDIA 13.6 — $0.20 $0.50 $0.28 423.9 0.53 — 49.5
GPT-6 Luna (Low) OpenAI 9.7 — $0.10 $0.50 $0.20 112.3 2.87 — 48.5
Granite 4.2 8B IBM 5.1 — $0.06 $0.25 $0.11 75.1 0.40 — 47.7
GPT-6 Luna (Non-reasoning) OpenAI 9.4 — $0.10 $0.50 $0.20 105.6 0.87 — 47.0
MiniMax-M3 MiniMax 18.9 — $0.30 $1.20 $0.53 101.6 0.63 — 36.0
DeepSeek V4.1 Flash 1017 (Non-reasoning) DeepSeek 18.5 — $0.30 $1.20 $0.53 157.0 0.97 — 35.2
Beam 1.0 Reflection 17.7 — $0.40 $1.20 $0.60 315.6 0.58 — 29.5
GPT-6.1 Terra (Max) OpenAI 51.3 — $1.00 $5.00 $2.00 109.8 242.32 — 25.6
Mistral Large 4 (High) Mistral 26.1 — $0.70 $2.10 $1.05 0.0 0.00 — 24.9
Inkling Small Thinking Machines 12.9 — $0.30 $1.20 $0.53 194.1 1.80 — 24.6
Mistral Large 4 Preview Mistral 25.3 — $0.70 $2.10 $1.05 0.0 0.00 — 24.1
GPT-6.1 Terra (Xhigh) OpenAI 45.3 — $1.00 $5.00 $2.00 96.7 53.42 — 22.6
Mistral Large 4 (Medium) Mistral 23.1 — $0.70 $2.10 $1.05 0.0 0.00 — 22.0
GPT-6.1 Terra (High) OpenAI 42.4 — $1.00 $5.00 $2.00 81.8 11.17 — 21.2
Gemini 3.8 Flash (High) Google 30.8 — $0.75 $3.75 $1.50 113.7 19.16 — 20.5
Gemini 3.8 Flash (Medium) Google 30.1 — $0.75 $3.75 $1.50 0.0 0.00 — 20.1
Step 5 Preview StepFun 28.5 — $1.00 $2.70 $1.43 133.4 0.77 — 20.0
GPT-6.1 Terra (Medium) OpenAI 38.2 — $1.00 $5.00 $2.00 86.0 3.05 — 19.1
Muse Spark 1.3 (Max) Meta 36.6 — $1.25 $4.25 $2.00 0.0 0.00 — 18.3
Qwen3.8 27B (Xhigh) Alibaba 20.1 — $0.50 $3.00 $1.13 38.0 1.39 — 17.9
Mistral Small 4 (Reasoning) Mistral 4.3 — $0.15 $0.60 $0.26 159.3 0.84 — 16.4
Qwen3.8 27B (Medium) Alibaba 17.5 — $0.50 $3.00 $1.13 43.0 1.26 — 15.6
GPT-6.1 Terra (Non-reasoning) OpenAI 30.5 — $1.00 $5.00 $2.00 80.6 0.94 — 15.3
Gemini 3.8 Flash (Low) Google 22.5 — $0.75 $3.75 $1.50 0.0 0.00 — 15.0
GLM-5.3 (Max) Z AI 31.7 — $1.40 $4.40 $2.15 63.2 1.63 — 14.7
GPT-6.1 Terra (Low) OpenAI 29.4 — $1.00 $5.00 $2.00 82.7 1.60 — 14.7
Qwen3.8 27B (Low) Alibaba 15.7 — $0.50 $3.00 $1.13 44.5 1.29 — 14.0
Gemini 3.5 Flash-Lite Google 11.7 — $0.30 $2.50 $0.85 362.5 5.73 — 13.8
GPT-6.1 Sol (Max) OpenAI 54.6 — $2.00 $10.00 $4.00 60.1 170.08 — 13.7
Claude Sonnet 5.5 (Max, Default Fallback) Anthropic 54.5 — $2.00 $10.00 $4.00 133.8 352.94 — 13.6
GPT-6.1 Sol (Xhigh) OpenAI 52.5 — $2.00 $10.00 $4.00 54.1 71.91 — 13.1
GPT-6.1 Sol (High) OpenAI 50.3 — $2.00 $10.00 $4.00 53.1 30.70 — 12.6
Claude Sonnet 5.5 (Xhigh, Default Fallback) Anthropic 49.9 — $2.00 $10.00 $4.00 92.5 16.02 — 12.5
Muse Spark Open (Xhigh) Meta 24.9 — $1.25 $4.25 $2.00 194.2 0.50 — 12.4
Muse Glimmer (High) Meta 6.3 — $0.30 $1.20 $0.53 111.3 0.43 — 12.0
Grok 4.7 (Xhigh) SpaceXAI 35.5 — $2.00 $6.00 $3.00 64.4 16.83 — 11.8
Grok 4.7 (High) SpaceXAI 35.2 — $2.00 $6.00 $3.00 60.7 15.07 — 11.7
DeepSeek V4 Pro 0813 (Max) DeepSeek 23.1 — $1.32 $3.96 $1.98 73.5 1.08 — 11.7
Kimi K3.1 Pro Kimi 46.5 — $2.00 $10.00 $4.00 28.4 2.30 — 11.6
GPT-6.1 Sol (Medium) OpenAI 46.3 — $2.00 $10.00 $4.00 52.0 4.78 — 11.6
Qwen3.8 Max (0902) Alibaba 33.7 — $2.00 $6.00 $3.00 37.0 1.97 — 11.2
GPT-6 Sol (Max) OpenAI 44.2 — $2.00 $10.00 $4.00 0.0 0.00 — 11.1
Gemini 3.1 Flash-Lite Google 6.2 — $0.25 $1.50 $0.56 0.0 0.00 — 11.0
GLM-5.3 (Low) Z AI 23.1 — $1.40 $4.40 $2.15 65.9 1.67 — 10.7
Nemotron 3 Ultra 550B A55B (Reasoning) NVIDIA 11.8 — $0.60 $2.70 $1.13 161.3 0.90 — 10.5
Claude Sonnet 5.5 (High, Default Fallback) Anthropic 41.5 — $2.00 $10.00 $4.00 91.7 8.79 — 10.4
Grok 4.7 (Low) SpaceXAI 31.1 — $2.00 $6.00 $3.00 57.1 2.29 — 10.4
Grok 4.6 (Xhigh) SpaceXAI 31.0 — $2.00 $6.00 $3.00 0.0 0.00 — 10.3
Qwen3.8 27B (Non-reasoning) Alibaba 10.0 — $0.50 $3.00 $1.13 43.6 1.27 — 8.9
GPT-6.1 Sol (Low) OpenAI 34.8 — $2.00 $10.00 $4.00 49.3 1.35 — 8.7
Claude Sonnet 5.5 (Medium, Default Fallback) Anthropic 30.8 — $2.00 $10.00 $4.00 87.3 0.89 — 7.7
Gemini 4 Carbon (High) Google 60.0 — $4.00 $20.00 $8.00 91.7 80.62 — 7.5
Claude Opus 5.5 (Max, Default Fallback) Anthropic 59.1 — $4.00 $20.00 $8.00 96.6 528.95 — 7.4
Gemini 4 Carbon (Medium) Google 58.3 — $4.00 $20.00 $8.00 75.7 31.08 — 7.3
GPT-5.6 Terra (Max) OpenAI 32.3 — $2.00 $12.00 $4.50 0.0 0.00 — 7.2
Claude Opus 5.5 (Xhigh, Default Fallback) Anthropic 57.4 — $4.00 $20.00 $8.00 77.2 69.87 — 7.2
Inkling (Xhigh) Thinking Machines 12.5 — $1.00 $4.05 $1.76 195.9 2.03 — 7.1
Gemini 4 Carbon (Low) Google 54.2 — $4.00 $20.00 $8.00 80.9 5.49 — 6.8
Claude Opus 5.5 (High, Default Fallback) Anthropic 51.8 — $4.00 $20.00 $8.00 69.3 19.40 — 6.5
Claude Sonnet 5.5 (Low, Default Fallback) Anthropic 25.6 — $2.00 $10.00 $4.00 88.6 0.94 — 6.4
DeepSeek V4 Pro 0813 (Non-reasoning) DeepSeek 12.6 — $1.32 $3.96 $1.98 77.4 1.04 — 6.4
Gemini 4 Argon (High) Google 50.4 — $4.00 $20.00 $8.00 84.6 32.94 — 6.3
Claude Opus 5.5 (Medium, Default Fallback) Anthropic 48.2 — $4.00 $20.00 $8.00 71.8 12.94 — 6.0
Muse Wonder (Max) Meta 43.6 — $4.00 $20.00 $8.00 125.1 30.14 — 5.5
Claude Opus 5.5 (Low, Default Fallback) Anthropic 34.8 — $4.00 $20.00 $8.00 68.9 6.25 — 4.3
Claude 4.5 Haiku (Reasoning) Anthropic 7.4 — $1.00 $5.00 $2.00 0.0 0.00 — 3.7
Kimi K3 (Low) Kimi 17.4 — $3.00 $15.00 $6.00 0.0 0.00 — 2.9
GPT-6 Astra (Max) OpenAI 56.9 — $10.00 $50.00 $20.00 48.7 234.63 — 2.8
GPT-6 Astra (Xhigh) OpenAI 55.3 — $10.00 $50.00 $20.00 47.0 103.67 — 2.8
GPT-6 Astra (High) OpenAI 54.8 — $10.00 $50.00 $20.00 41.3 44.06 — 2.7
GPT-6 Astra (Medium) OpenAI 51.6 — $10.00 $50.00 $20.00 44.7 4.68 — 2.6
Claude Fable 5.1 (Xhigh, Default Fallback) Anthropic 50.0 — $10.00 $50.00 $20.00 51.0 34.21 — 2.5
Claude Fable 5.1 (Max, Default Fallback) Anthropic 49.2 — $10.00 $50.00 $20.00 69.2 177.05 — 2.5
Claude Fable 5.1 (High, Default Fallback) Anthropic 47.2 — $10.00 $50.00 $20.00 50.4 8.43 — 2.4
GPT-6 Astra (Low) OpenAI 43.3 — $10.00 $50.00 $20.00 40.9 1.81 — 2.2
Claude Fable 5.1 (Medium, Default Fallback) Anthropic 43.2 — $10.00 $50.00 $20.00 47.6 3.72 — 2.2
Claude Fable 5.1 (Low, Default Fallback) Anthropic 38.7 — $10.00 $50.00 $20.00 47.5 0.76 — 1.9
Mistral Medium 3.5 Mistral 4.6 — $1.50 $7.50 $3.00 168.4 0.83 — 1.5
Muse Wonder Full (Max) Meta 50.1 — $0.00 $0.00 $0.00 0.0 0.00 — 0.0
Nanbeige4.3-33B Nanbeige 20.4 — $0.00 $0.00 $0.00 0.0 0.00 — 0.0
K2 Horizon 375B A23B Institute of Foundation Models 17.8 — $0.00 $0.00 $0.00 116.8 2.23 — 0.0
Command A+ Cohere 6.2 — $0.00 $0.00 $0.00 209.1 0.19 — 0.0
Granite 4.2 3B IBM 4.7 — $0.00 $0.00 $0.00 0.0 0.00 — 0.0
NVIDIA Nemotron 3 Nano 4B NVIDIA 1.8 — $0.00 $0.00 $0.00 0.0 0.00 — 0.0

The most expensive model is rarely the best choice

There is a persistent assumption in the industry that higher price means higher quality. The data tells a different story. Sort the table above by value score and you will see that many of the top-ranked models cost a fraction of the flagship offerings while matching or exceeding them on intelligence and coding benchmarks. A model at $0.50 per million tokens can score within a few points of one at $15 — that is a 30x price difference for nearly identical capability.

This matters at scale. If your application processes a million requests a day, the difference between a $2/M model and a $0.20/M model is not a rounding error — it is thousands of dollars a month. And the cheaper model might actually respond faster, because smaller or better-optimized models frequently achieve higher throughput. You are paying more for less speed.

The flagship models from the largest providers do lead on the hardest benchmarks — frontier math, PhD-level science, complex multi-step reasoning. But most production workloads are not PhD-level science. They are classification, extraction, summarization, code generation, and conversational tasks. For these, a mid-tier model is not a compromise — it is the right tool. Using a $10/M model to parse invoices is like hiring a surgeon to apply a bandage.

Reasoning models add another dimension. They spend extra compute “thinking” before answering, which boosts accuracy on hard problems but also increases latency and cost. If your use case does not require multi-step logical deduction, a standard model will give you the same answer in a tenth of the time. The filter above lets you isolate reasoning models so you can compare them separately.

The value score in this table — intelligence divided by blended price — exists to make this tradeoff visible at a glance. The best model for your project is not the smartest one available. It is the smartest one you need, at the lowest price that delivers it.

If you are building on European infrastructure and want to find providers that respect data sovereignty, check out Voie.fi — an open index of 1,400+ European-headquartered digital infrastructure providers across compute, storage, payments, security, AI, and more. It helps teams replace US-dominant services with genuinely European alternatives that avoid CLOUD Act and FISA exposure.