Alibaba (Qwen)

Qwen2 72B Instruct

Type
language model
Context
33K tokens
Max output
33K
API string
fireworks_ai/accounts/fireworks/models/qwen2-72b-instruct

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
Fireworks AI $0.9 $0.9 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, mathematics 1,234.83 1,225–1,244
Arena Score in English 1,222.22 1,216–1,228
Arena Score, overall 1,203.43 1,199–1,208
Arena Score, programming 1,196.31 1,187–1,205
Arena Score, multi-turn dialogue 1,195.55 1,186–1,205
Arena Score, long queries 1,192.58 1,182–1,203
Arena Score, hard prompts 1,190.89 1,184–1,198
Arena Score, instruction following 1,181.62 1,175–1,188
Arena Score, creative writing 1,181.3 1,172–1,191
Arena Score, expert questions 1,170.74 1,156–1,185
Arena Score in French 1,169.13 1,143–1,196
Arena Score in Spanish 1,168.13 1,140–1,196
Arena Score in Russian 1,167.83 1,157–1,178
MATH, difficulty level five 39.07 % · with a tuned harness Epoch evaluations
GPQA Diamond — graduate-level questions 21.04 % · with a tuned harness Epoch evaluations
WeirdML — unusual machine learning tasks 11.3 % https://htihle.github.io/weirdml.html
The Agent Company — work tasks in an office environment 1.1 % TheAgentCompany experiment results github