ModelPriceWatch.com
Last scan 2026-11-07 Models tracked 308 Providers 42 Cheapest paid pplx-decider-flash-v1.2 $0.010/Mtok in Every price links to its source

Mistral Large 3

by Mistral

Deprecated mid tier open weights cheap tier
Email me when Mistral Large 3 gets cheaper
Deprecated — shuts down . Still purchasable until then, but the provider has announced end-of-life. Don't start new work on it. Migration target: Mistral Large 4. Full deprecation status & replacements →
Availability: Mistral deprecated Mistral Large 3 (mistral-large-2512) on 2026-10-31 and will retire it on 2027-04-30, naming Mistral Large 4 as the replacement. It is no longer on Mistral's API price list, but the API still serves it, and Mistral's own endpoint on OpenRouter still charges $0.5 input / $1.5 output per 1M tokens, the price shown. Read 2026-11-02.

Today's price · per 1M tokens

Input

$0.500

per 1M tokens

Output

$1.50

per 1M tokens

Blended

$0.750

blended $/1M — 3:1 weighted input:output

Source: official Mistral pricing · read MODELPRICEWATCH.COM · 2026-11-07

Price receipt

We read Mistral's own pricing page on and found Mistral Large 3 ↗ listed with a price on the same row — that read is where the number above comes from. We keep the page text we read; its content hash is e04eb8c2f5.

Overview

Mistral's previous large open-weight model, a 675B-parameter MoE (41B active) under Apache 2.0, strong on knowledge and math (93.6% MATH-500), though not a dedicated reasoning model. Deprecated in favour of Mistral Large 4. $0.50/$1.50 per 1M tokens.

Run it yourself

Deploy this open model on rented GPUs

Open weights mean you can self-host instead of paying per-token API prices. These platforms let you serve it on demand — often cheaper at scale.

RunPod →
Rent GPUs by the second and serve any open-weight model yourself.
Novita AI →
Serverless open-model inference billed per token — no GPU ops.
Vast.ai →
GPU marketplace — rent spot/on-demand GPUs cheaply and serve any open-weight model.

Partner links — we may earn a commission if you sign up, at no cost to you. We only list platforms we'd recommend regardless.

Capabilities

struck through = not supported
Input 1/4
Text ✓ Image Audio Video
Output
Text ✓
Features 2/9
Prompt caching Reasoning Coding Fast inference Long context ✓ Open weights ✓ Multimodal Web search Realtime

Benchmark performance

accuracy % · higher is better
Avg benchmark score53.6%
Perf per $/Mtok71.5
GPQA Diamond
58%
SWE-Bench Verified
35%
Humanity's Last Exam
18%
ARC-AGI 2
20%
AIME 2025
52%
MMMLU
78%
BFCL
68%
HumanEval
85%
MATH 500
68%
Independent composite scores
  • AA Intelligence Index: 9.3 (Artificial Analysis composite across reasoning, knowledge and coding evals)
  • AA-Omniscience Index: -39.6 (−100–100)
  • GDPval-AA v2: 427.6 ELO (Real-world work tasks, human baseline = 1000)
  • LMSYS Chatbot Arena: 1414.4 ELO (Human preference)
How it stacks up
  • Ranks #77 of 100 benchmarked models by average score
  • Ranks #123 of 194 comparably-measured models by percentile score, across 13 independent measurements
  • Ranks #72 of 194 comparably-tested models by normalized performance per dollar
  • Strongest at HumanEval — 85%, #9 of 47

120 tokens/sec output

Source: Artificial Analysis, Mistral model card, Artificial Analysis, LMSYS Chatbot Arena (UC Berkeley) · updated · See full rankings →

Specifications

Provider
Mistral
Context window
262K tokens
Modality
text
Parameters
Proprietary
Open source
Yes — open weights available
Released
Status
Deprecated ends
Last updated
Tags
mid-tier

Availability verified: — per Mistral's own deprecation notice

Try on Mistral Official pricing page