Alibaba (Qwen)

Qwen3 Max 2025-09-23

Type
language model
Context
258K tokens
Max output
66K
Released
24 September 2025
API string
qwen3-max-2025-09-23

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
302.AI $0.86 $3.43 31 Jul 2026

Only the base tier and only per-token prices are shown. Batch, discounted and cached rates, as well as prices per image or per second of video, do not go into this table: they cannot stand in the same column as a price per million tokens.

The date in the last column is the day this price first entered our collection. The price may well be older: before that day we simply were not recording it. It has not changed since — otherwise a new row with a new date would stand in its place.

Measurement results

Task set Result Run conditions Measured by
Arena Score, programming 1,439.19 1,426–1,452
Arena Score, mathematics 1,438.73 1,415–1,462
Arena Score, multi-turn dialogue 1,436.77 1,422–1,451
Arena Score, hard prompts 1,424.87 1,416–1,433
Arena Score in English 1,420.37 1,412–1,429
Arena Score in Russian 1,419.24 1,396–1,443
Arena Score, overall 1,413.19 1,407–1,420
Arena Score, long queries 1,412.27 1,400–1,425
Arena Score, instruction following 1,399.35 1,388–1,411
Arena Score, expert questions 1,397.86 1,371–1,425
Arena Score in Spanish 1,397.75 1,368–1,428
Arena Score, creative writing 1,382.56 1,366–1,400
MATH, difficulty level five 97.13 % · with a tuned harness Epoch evaluations
Mock AIME 2024–2025 — olympiad problems 73.31 % · with a tuned harness Epoch evaluations
SimpleQA Verified — factual accuracy 67.47 % Epoch evaluations
Fiction.LiveBench — holding a long context 66.7 % Fiction.live leaderboard
GPQA Diamond — graduate-level questions 63.47 % · with a tuned harness Epoch evaluations
Chess puzzles 0 % Epoch evaluations