Skip to content
OSMEXO
Sign inGet an API key
☰
Sign inGet an API key

Qwen3.8 27B

alibaba/qwen3.8-27b
  • Alibaba
  • Family · Qwen3
  • Released 14 Aug 2026
  • Open weights
  • Text · Image · Video → Text
  • Reasoning · adjustable
  • Tools · JSON · Vision
Not on OSMEXO yet
Request this model
Spec & list price
Context1M
Max output236K
Input / 1M$0.15
Output / 1M$1.8812.5× input
Scores · standing
Intelligence33.7#54/489
LiveBench58.4#25/32
Coding68.1#32/196
Usage
Tokens / day126.1B▲ 6.1% 7d
Usage rank#30
Requests / day4.4M
Tool calls / req0.26
Per request · 30 days
Avg input16.6K1.1× median
Avg output717about median
Cache hit73%about median
Reasoning share66%1.2× median
Tool-call failures1.4%0.7× median

Consumption

40.0B80.0B120.0B3 Oct17 Oct1 Nov
Peak 102.0B 4 OctLow 37.5B 27 SepRank today #30

News & trends

36 mentions this week ▼ 5%this model · videos, posts, articles · 30 days
Trending · 14 days
News

Benchmarks

bar = share of the top score · Benchmark data: Artificial Analysis
GPQA Diamond#36 of 445
90.5%
τ-Bench Banking#6 of 154
48.0%
Long-context reasoning#33 of 393
82.0%
Terminal-Bench 2.1#24 of 182
79.8%
Coding index#32 of 196
68.1
SciCode#77 of 160
46.6%
Intelligence index#54 of 489
33.7
Humanity's Last Exam#79 of 465
33.9%
Speed · median across 22 providers, latest day
Output speed#243/33448 tok/s1.4× slower than median
Fastest provider663 tok/sbest of 22 providers
First token#1470.92 s1.1× quicker than median
First answer token#8850 sthinking included

#2 cheapest of 9 with a similar score

Value
Intelligence Index
010203040$0.1$0.2$0.5$1$2$5Qwen3.8 27B
Blended price per 1M tokens (log)
Family progress
Intelligence Index
010203040MayJulSepNov2026MarMayJulQwen3.8 27B
  • Family releases33.7
LiveBench suites
Browser24.2
Coding86.8
Data Analysis76.6
Design75.9
Long Agentic20.6
Reasoning80.0
Research59.4

▲ Strongest: Coding, Reasoning. ▼ Weakest: Long Agentic (21).

Technical details

Reasoning
on · xhigh
Tokenizer
Qwen
License
Apache 2.0
Parameters
27.8B
Moderation
none
Downloads · likes
7.9M · 19.0K
Reasoning effort
  • xhigh · default
  • medium
  • low
  • none
Request parameters 21
  • tools
  • tool_choice
  • response_format
  • structured_outputs
  • reasoning
  • reasoning_effort
  • include_reasoning
  • temperature
  • top_p
  • top_k
+11 more
  • min_p
  • max_tokens
  • stop
  • seed
  • frequency_penalty
  • logit_bias
  • logprobs
  • presence_penalty
  • repetition_penalty
  • top_a
  • top_logprobs

Interesting facts

  1. 1.8× the output room of similar models (235.9K against a 129.5K median).
  2. The most used of 15 Qwen3 models by tokens last month.
  3. Released 80 days ago, the newest of 25 Qwen3 models.

When to use

Best for

  • Open-weight coding assistants
  • Reasoning and science questions
  • Image and video understanding
  • Long document question answering

Think twice for

  • Long autonomous agent runs (weakest suite)
  • Browser automation
  • Deep research tasks

Research papers

Same family

Qwen3

Alternatives

other vendors · similar capability

Want Qwen3.8 27B on OSMEXO?

Sign in to request it. Requests are reviewed; a request does not guarantee availability.

Request this model on OSMEXO