invokeable()
← comparisons / gpt-5.6-luna-vs-mistral-medium-2604

GPT-5.6 Luna vs Mistral Medium 3.5

Side-by-side from the Invokeable directory — provider-attributed pricing, canonical IDs, draft data flagged. Open as a shareable card →

01Head to head

GPT-5.6 LunaMistral Medium 3.5
CreatorOpenAIMistral AI
Released2026-07-092026-04-29
Context1.1M262K
Max output128K262K
Input $/M——
Output $/M——
Knowledge cutoff2026-02—
Modalitiestext + image → texttext + image → text
Capabilitiestool-use, extended-thinking, structured-outputtool-use, extended-thinking, structured-output
ProvidersOpenAI APIMistral AI (La Plateforme)

02Benchmarks · vendor-reported unless verified

GPT-5.6 LunaMistral Medium 3.5
BrowseComp (accuracy)83.3—
DeepSWE67.2%—
FrontierMath (accuracy)78.6—
GPQA Diamond (accuracy)92.3—
SWE-bench Pro62.7%—
SWE-bench Verified (resolved)—77.6
Terminal-Bench (success rate)84.7—
τ³-Telecom (accuracy)—91.4