Home  /  Compare  /  GPT-6 Luna vs Mistral Large 4
Head to head

GPT-6 Luna vs Mistral Large 4.

Verdict by category, tale of the tape, shared benchmarks only, and cost at three workloads — all from the tracked data.

VS
Mistral AI

Verdict by category

CategoryGPT-6 LunaMistral Large 4Winner
League Score53.353.1GPT-6 Luna
Coding (SWE-bench %)———
Reasoning (GPQA %)———
Context (tokens)10500001000000GPT-6 Luna
Speed (tok/s)———
Price (blended $/1M)$0.20$1.05GPT-6 Luna (cheaper)

Tale of the tape

SpecGPT-6 LunaMistral Large 4
MakerOpenAIMistral AI
ReleasedSep 22, 2026Oct 6, 2026
League Score5353
Context1,050,000 tokens1,000,000 tokens
Max output128,000—
$/1M input$0.10$0.70
$/1M output$0.50$2.10
WeightsClosedOpen
Speed——

Head-to-head benchmarks (shared only)

BenchmarkGPT-6 LunaMistral Large 4
LiveBench overall60.460.3

Cost at three workloads

WorkloadGPT-6 LunaMistral Large 4
Light — 1M in / 250K out$0.23$1.23
Standard — 10M in / 2.5M out$2.25$12.25
Long-context — 50M in / 5M out$7.50$45.50

The pick

Pick GPT-6 Luna if you want the higher overall League Score; you need the bigger context window; you want the lower running cost.

Pick Mistral Large 4 if you want downloadable weights / self-hosting.

All figures from the model database — sources and dates on each model page. Cost your own workload or compare a different pair.

Frequently asked questions

Which is better overall, GPT-6 Luna or Mistral Large 4?

On overall League Score they are effectively tied — GPT-6 Luna scores 53 and Mistral Large 4 scores 53. The category verdicts on this page are where the real differences live.

Which is cheaper, GPT-6 Luna or Mistral Large 4?

GPT-6 Luna is cheaper: $0.20 vs $1.05 blended per 1M tokens (3:1 input/output mix) — about 5.2x less per token.

Which has the bigger context window, GPT-6 Luna or Mistral Large 4?

GPT-6 Luna, with 1.1M of context against 1M for Mistral Large 4.

Are GPT-6 Luna and Mistral Large 4 open source?

GPT-6 Luna is closed weights — API access only; Mistral Large 4 is open weights — it can be downloaded and self-hosted.