OpenRating is an independent, open-source AI rating platform. Join the community

OOpenRating

Coding workloads

Best LLM for coding

Ranked on the axes OpenRating measures — reasoning capability, cost per task, and output throughput — across all 15 rated models. Claude Opus 5 (max) leads on reasoning at 59.4.

Read this first

OpenRating does not yet run a coding-specific benchmark. Every figure below comes from the same standardised workload we apply to all models — reasoning, knowledge, and agentic tasks — so this ranking reflects measured reasoning capability and measured economics, not a coding score. We publish it this way rather than invent a metric we have not earned.

The ten strongest reasoning models

Reasoning capability is the closest thing we measure to coding ability. Cost and speed decide whether it is usable at scale.

#ModelProviderReasoningCost / taskSpeedContextLicense
1Claude Opus 5 (max)Anthropic59.4$11.8279 tok/s1MProprietary
2GPT-6 Sol (Max)OpenAI59.4$2.47122 tok/s1MProprietary
3Grok 4.6 (high)xAI41.9$4.6878 tok/s500KProprietary
4GPT-5.6 Luna (Max)OpenAI32.4$0.3399 tok/s1MProprietary
5Kimi K3 (Low)Moonshot AI23.0$2.3143 tok/s1MProprietary
6Gemini 3.5 Flash-LiteGoogle16.2$0.69311 tok/s1MProprietary
7Inkling (Xhigh) (Xhigh)Inkling16.2$1.3595 tok/s1MOpen weights
8Nemotron 3 UltraNVIDIA16.2$2.19218 tok/s262KOpen weights
9Claude 4.5 HaikuAnthropic9.5$0.5391 tok/s200KProprietary
10Nemotron 3.5 LightningNVIDIA8.1$0.31273 tok/s1MOpen weights

Why cost per task beats price per token

Coding is the workload where token pricing misleads most. An agent working through a repository does not send one prompt and stop — it reads files, writes patches, runs tests, reads the failures, and tries again. Each loop resends context, so a model that looks cheap per million tokens can be the expensive one by the end of a task.

Cost per task measures the whole workload instead: reasoning tokens, answer generation, cache writes, cache hits, and input payload, summed for one standardised task. That is the number that scales with your agent's retry behaviour.

Reasoning tokens are usually the largest line. A model that thinks longer is not being wasteful if it needs fewer attempts.

Estimate your own token costs →

Best reasoning per dollar

Reasoning index divided by task cost. A derived ratio from two measured values — not a benchmark.

GPT-5.6 Luna (Max)

OpenAI · $0.33 per task

98.2

Muse Glimmer (high)

Muse AI · $0.18 per task

45.0

Nemotron 3.5 Lightning

NVIDIA · $0.31 per task

26.1

GPT-6 Sol (Max)

OpenAI · $2.47 per task

24.0

Gemini 3.5 Flash-Lite

Google · $0.69 per task

23.5

If your code cannot leave the building

For many teams the deciding constraint is not capability or cost — it is whether proprietary source can be sent to a hosted API at all. Open-weights models can be self-hosted, which removes that question entirely and moves the cost from per-task billing to fixed infrastructure. The trade is real: you take on serving, quantisation, and upgrade work that a hosted API absorbs for you.

How this ranking is built

Every model is scored on the same standardised workload. Reasoning index is a 0–100 composite of multi-step reasoning tasks. Cost per task is the full end-to-end spend for one benchmark task, including reasoning tokens, cache writes and hits, answer generation, and input payload. Speed is blended output generation in tokens per second under standardised load. Nothing on this page is a vendor-reported figure.