Mistral Large 3 is a MoE model with 67.5B total parameters and 41B active parameters, supporting a 256K-token context window. Trained from scratch on 3,000 NVIDIA H200 GPUs, it is one of the strongest permissively licensed open-weight models available. Designed for advanced reasoning and long-context understanding, Mistral Large 3 delivers performance on par with the best instruction-tuned open-weight models for general-purpose tasks, while also offering image understanding capabilities. Its multilingual strengths are particularly notable for non-English/Chinese languages, making it well-suited for global applications. Typical use cases include enterprise assistants, multilingual customer support, content generation and editing, data analysis over long documents, code assistance, and research workflows that require handling large corpora or complex instructions. With its MoE architecture, Mistral Large 3 balances strong performance with efficient inference, providing a versatile backbone for building reliable, production-grade AI systems.
Mistral Large 3 vs Qwen3.8 Max Preview
Output tokens cost $1.50 per million on Mistral Large 3 and $1.01 per million on Qwen3.8 Max Preview. Input tokens cost $0.50 per million on Mistral Large 3 and $0.34 per million on Qwen3.8 Max Preview. Cached input tokens are billed at $0.50 per million on Mistral Large 3 and $0.03 per million on Qwen3.8 Max Preview. Context lengths are 256,000 tokens on Mistral Large 3 and 983,616 tokens on Qwen3.8 Max Preview. Time to first token (TTFT) measured on AIHubMix is 1.9s on Mistral Large 3 and 2.4s on Qwen3.8 Max Preview. Measured output throughput is 51.6 tok/s on Mistral Large 3 and 48.1 tok/s on Qwen3.8 Max Preview. On the LMArena coding leaderboard Mistral Large 3 scores 1468 and Qwen3.8 Max Preview scores 1520.
Qwen 3.8 Max Preview(Qwen3.8-Max-Preview) is the latest-generation foundation model in the Qwen family, packing 2.4T parameters and still evolving. Compared with the previous flagship Qwen 3.7 Max, it delivers major gains in core capabilities like Coding and Cowork (professional productivity), with world-leading performance on complex, long-horizon tasks such as full-stack development, data analysis, and Office workflows. Launch offer: Credits are consumed at just 20% of the standard rate, effectively 5× your usage. Limited time only.
Pricing & Specifications
Prices are per million tokens. Time to First Token and throughput are rolling averages measured on AIHubMix.
Promotional prices show the discounted rate; see each model page for promotion windows.
Activity Past 30 Days
Daily traffic served through AIHubMix — how demand for each model is trending.
Tokens / day
Requests / day
Performance Past 3 Days
Measured on real AIHubMix traffic, hourly buckets. Gaps mean no traffic in that hour.
Throughput (tok/s)
TTFT (s)
Uptime (%)
LMArena Benchmarks
LMArena ratings by capability (Bradley-Terry, commonly called Elo). Higher is better.
Source: LMArena (arena.ai) leaderboard, imported by AIHubMix. Models without published ratings are omitted per chart.
Cost calculator
Estimate your monthly bill for the same workload on each model.
Monthly = daily × 30. Discounted rates applied where a promotion is active.
FAQ
Which is cheaper: Mistral Large 3, Qwen3.8 Max Preview?
Qwen3.8 Max Preview: $1.01/M output tokens; Mistral Large 3: $1.50/M. Use the cost calculator above to estimate your own workload.
How do their coding arena scores compare?
Qwen3.8 Max Preview: 1520; Mistral Large 3: 1468 (LMArena coding leaderboard).
Which responds faster?
Mistral Large 3: 1.9s time to first token measured on AIHubMix; see the live performance charts above for how each model behaves across the day.
How large is each context window?
Mistral Large 3 accepts 256,000 and Qwen3.8 Max Preview accepts 983,616 input tokens. Maximum output per request is 256,000 tokens on Mistral Large 3 and 131,072 tokens on Qwen3.8 Max Preview.
Which one generates tokens faster?
Mistral Large 3 at 51.6 tok/s and Qwen3.8 Max Preview at 48.1 tok/s, measured as output throughput on AIHubMix — a separate metric from time to first token.
What inputs and capabilities does each model support?
Mistral Large 3 accepts text and image input and supports function calling and structured outputs; Qwen3.8 Max Preview accepts text and image input and supports tool calling, function calling, structured outputs, web search, long context and thinking.
Can I call Mistral Large 3 and Qwen3.8 Max Preview with the same API key?
Yes. AIHubMix serves every model on this page behind one OpenAI-compatible endpoint, so switching between them is a one-line change to the model field — no second account, key or SDK.
Popular comparisons
Related model match-ups readers also look at.
