Thread

Respond to this Pin

Terminal Bench 2.1%

Qwen3.7-Max74.5
Opus 4.884.6
Fable 584.6
GPT-5.6 Sol (max)88.8
Qwen3.8-Max86.6

Alibaba's model-card table; harnesses differ by model (Qwen on Claude Code avg@10, others best published across harnesses). · source

SWE-bench Pro%

Qwen3.7-Max60.6
GPT-5.6 Sol (max)64.6
Opus 4.869.2
Fable 580
Qwen3.8-Max67.7

Alibaba's model-card table, Claude Code harness on a corrected task set. · source

DeepSWE 1.1%

Qwen3.7-Max21.6
Qwen3.8-Max56.6
Opus 4.859
Fable 570
GPT-5.6 Sol (max)73

Alibaba's model-card table; best of Claude Code and mini-SWE-agent harnesses. · source

HLE with tools%

Qwen3.7-Max53.5
Qwen3.8-Max56.2
Opus 4.857.9
GPT-5.6 Sol (max)58
Fable 564.5

Alibaba's model-card table. · source

Qwen3.8-Max specifications

Item	Value	Basis
Parameters	2.4T total, 95B activated, 92 layers, 512 experts (10 routed + 1 shared)

Language model rubric2026 criteria

Estimate from the version's published specs, benchmarks and reviews, 0-5; scores are only comparable within one criteria year. Scored as launched (3 Aug 2026; weights promised the following week). Reasoning: AA Intelligence Index 56 at launch on v4.1 (level with Claude Opus 4.8, behind Kimi K3 57; The Decoder, OfficeChai, NYU Shanghai RITS 6-7 Aug); AA's first run of 53 was withdrawn after endpoint issues and RITS says it was later revised to 58. AA's current page shows 40 on v4.3.2, ignored. Coding null: no TB 4.0; Alibaba's own Terminal-Bench 2.1 is 86.6 (not mapped). Context 1M tokens. Price $2.00 in / $6.00 out per 1M on Alibaba Cloud = $3.00 blended (down from $2.50/$7.50 for 3.7 Max); per-task cost rose because it needs 64 turns instead of 14. Speed 42 tok/s as measured by AA (page now). Modalities text, image, video in, text out. CNBC piece only covers the launch.

User Photos

Photos from readers, kept apart from the pin's sourced media. Rate what you ate or saw, and on a restaurant pin tag the dishes right on your photo.

Comments

References 781% CONFIDENCEOverall confidence: 81%How well the pin's source and references back up its dates.Weighted average of how firmly 7 references, the source included, support the pin's start and end times; a reference counts half as much for every 180 days older than the newestShow all pins at 75% confidence or better

The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.

[1]90%
alibabagroup.com/en-US/document-2021044032125272064
alibabagroup.com· Posted Sep 28, 2026· Starts Aug 3, 2026 ✓· 16% of score
Alibaba Group's release is dated "August 3, 2026" and says "The model is now accessible via APIs on the Alibaba Cloud Model Studio for global developers, with model weights scheduled for release next week"; CNBC^[6] reports Alibaba "unveiled" it "on Monday" 3 August.
[2]90%
qwen3.8-max Model Info - Alibaba Cloud Model Studio
help.aliyun.com· Added Sep 28, 2026· 16% of score
Alibaba Cloud's model page: model ID qwen3.8-max, now the snapshot qwen3.8-max-0902 (alias qwen3.8-max-2026-09-02); image, text and video input; a 1,000,000-token context window with 131,072 max output; function calling, structured outputs, web search and context caching; CNY 12 input and CNY 36 output per million tokens in Beijing.^[1]
[3]75%
Qwen - Wikipedia
en.wikipedia.org· Added Sep 28, 2026· 16% of score
Dates the cloud release to 3 August 2026, the open weights (Qwen3.8-2.4T-A95B, without image input or non-thinking mode) to 12 August, with a licence requiring providers earning more than US$50 million to get a commercial licence, and the Apache 2.0 Qwen3.8-27B to 14 August.^[1]
[4]70%
Qwen3.8 Max Prime: Is it actually better? - DataNorth AI
datanorth.ai· Published Sep 24, 2026· 15% of score
Reports that Alibaba's Qwen team released Qwen3.8 Max Prime on 23 September 2026 as a higher-throughput version of Qwen3.8 Max at $4 per million input and $12 per million output tokens, "exactly twice the price of the standard model", and that it was not measurably faster in OpenRouter's first measurements.^[1]
[5]90%
Qwen/Qwen3.8-2.4T-A95B · Hugging Face
huggingface.co· Published Aug 12, 2026· 13% of score
The open-weights model card: "For the first time, Qwen3.8 brings a Qwen-Max-class model to open release"; 2.4T total and 95B activated parameters, 92 layers, 512 experts (10 routed + 1 shared), 262,144 tokens natively extensible to 1,010,000, and benchmarks against Opus 4.8, Fable 5 and GPT-5.6 Sol.^[1]

Suggest a correction

Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most. A picture that shows something else, or shows it badly, is looked at too, and moved down or replaced.

Reviews

Qwen3.8 Max catches Claude Opus 4.8 but Kimi K3 still scores higher for 25 percent less · the-decoder.com · Aug 6, 2026Qwen3.8 Max needs 64 steps per task instead of 14, and input tokens grew 15x because the test resends the full conversation history to the model at each step.
Qwen 3.8 Max Scores 56 On Artificial Analysis Intelligence Index, Ahead Of All US Companies Except Anthropic And OpenAI · officechai.com · Aug 6, 2026Qwen3.8 Max is now attempting questions it previously would have declined to answer, and getting a lot of them wrong instead.
Qwen3.8-Max Draws Level With the Frontier on Agentic Benchmarks · rits.shanghai.nyu.edu · Aug 7, 2026On Artificial Analysis' Agentic Index , the 2.4-trillion-parameter model scores 58 — tied with Claude Opus 5 running at xhigh effort, and one point behind Opus 5 at max effort

Sentiment over timeMixed

Sentiment

Sentiment over time
Qwen3.8 Max catches Claude Opus 4.8 but Kimi K3 still scores higher for 25 percent less · Aug 6, 2026	-0.50
Qwen 3.8 Max Scores 56 On Artificial Analysis Intelligence Index, Ahead Of All US Companies Except Anthropic And OpenAI · Aug 6, 2026	-0.40
Qwen3.8-Max Draws Level With the Frontier on Agentic Benchmarks · Aug 7, 2026	0.60

Suggest a correction

Comments

Reviews

Sentiment over timeMixed

Qwen3.8-Max Released

Notable features

Tags

Thread