geekradar
Int
Explore GeekRadar

More to explore

Good gear All topics About us Say hello
GeekRadar / The comparison desk
COMPARE.
AI models
alibaba

Qwen3.8 Max (0902)

AI model Released

Context window
1,000,000 tokens*
Maximum output
131,072 tokens*
Open weights
No*

Capability & preference

General capability · ECI

A cross-benchmark capability estimate. Index points, not a percentage; intervals show uncertainty.

Qwen3.8 Max (0902) 155.9* Qwen3.8 Max (0902) · 95% interval 153.7–158.1

Model performance

Radar task score: four tests, equal weight. An experimental summary of these tasks, not a universal intelligence rating.

Qwen3.8 Max (0902)

Radar task score

Unrated

3/4 core tests · not enough evidence

7 benchmark results for this configuration

Scientific reasoning

GPQA Diamond · 25%

Expert-written science questions; strict multiple-choice scoring. Chance accuracy is 25%.

Qwen3.8 Max (0902) 92.3*% Approx. 95% interval: 88.9–95.7%

Novel problem solving

Mystery Game Puzzles · 25%

Best-next-move puzzles in an undisclosed game variant. A narrow reasoning test, not general intelligence.

Qwen3.8 Max (0902) Not available

Advanced mathematics

FrontierMath Tiers 1-3 · 25%

Private v2 mathematics problems with Python tools. Earlier versions and Tier 4 are excluded.

Qwen3.8 Max (0902) 65.6*% Approx. 95% interval: 60.1–71.1%

Factual answers

SimpleQA Verified · 25%

Correct factual answers under a guessing prompt, graded by a model. Not a hallucination or abstention score.

Qwen3.8 Max (0902) 47.3*% Approx. 95% interval: 44.2–50.4%

End-to-end software

MirrorCode · Separate result

Whole-program reimplementation with hidden tests; ML + Private, two languages. Budget: seven days and ten billion tokens per attempt.

Qwen3.8 Max (0902) 47.8*% Approx. 95% interval: 27.6–67.9%

Competition mathematics

OTIS Mock AIME 2024-2025 · Separate result

Competition problems with model-extracted answers checked against an exact key. Earlier grading procedures are excluded.

Qwen3.8 Max (0902) 100.0*% Uncertainty not available

Research mathematics

FrontierMath Tier 4 v2 · Separate result

The hardest private FrontierMath problems, corrected v2 set, with Python tools. Separate from Tiers 1-3 in the core score.

Qwen3.8 Max (0902) 34.1*% Approx. 95% interval: 19.5–48.8%

Visual spatial reasoning

Furniture Assembly · Separate result

Assembly-error detection from photos and manuals, with zoom and Python tools. Uses model-assisted grading; not a general vision score.

Qwen3.8 Max (0902) 20.0*% Approx. 95% interval: 9.8–30.2%

Bars show reported task scores. Whiskers are approximate 95% intervals where standard errors are available. Different reasoning budgets are not equal-cost tests. Missing results are never zero.

Run it locally

Inference · memory estimate

Qwen3.8 Max (0902)

Closed weights. No supported local deployment or GPU requirement is published.

Capabilities

Context window
1,000,000 tokens*
Maximum output
131,072 tokens*
Reasoning
Yes*
Tool calling
Yes*
Open weights
No*

Input modalities

  • text*
  • image*
  • video*
  • pdf*

Output modalities

  • text*

System configuration

Detailed specifications

Specifications 1 · 4 fields
Access model
API access*
Open weights
No*
Organization
Alibaba*
Organization country
China*
Specifications 2 · 19 fields
File attachments
Yes*
Context window
1,000,000* tokens
Model family
qwen*
Input modalities
  • text*
  • image*
  • video*
  • pdf*
Maximum output
131,072* tokens
Output modalities
  • text*
API service · Ofox · qwen/qwen3.8-max-0902
Provider
Ofox*
Model
qwen/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
  • pdf*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    toggle*
  • type
    effort*
    values
    • low*
    • medium*
    • xhigh*
  • type
    budget_tokens*
    min
    0*
    max
    262,144*
Tool calling
Yes*
Structured output
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
1.71*
output
5.14*
cache_read
0.17*
cache_write
2.14*
API service · EmpirioLabs AI · qwen3-8-max-0902
Provider
EmpirioLabs AI*
Model
qwen3-8-max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    effort*
    values
    • none*
    • low*
    • medium*
    • high*
    • max*
  • type
    budget_tokens*
    min
    1*
    max
    262,144*
Tool calling
Yes*
Structured output
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
2*
API service · Ofox · bailian/qwen3.8-max-0902
Provider
Ofox*
Model
bailian/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
  • pdf*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    toggle*
  • type
    effort*
    values
    • low*
    • medium*
    • xhigh*
  • type
    budget_tokens*
    min
    0*
    max
    262,144*
Tool calling
Yes*
Structured output
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
cache_write
2.5*
API service · NanoGPT · qwen/qwen3.8-max-0902
Provider
NanoGPT*
Model
qwen/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
991,808*
input
991,808*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
  • pdf*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    effort*
    values
    • none*
    • low*
    • medium*
    • high*
    • max*
Tool calling
Yes*
Structured output
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
cache_write
2.5*
API service · CommandCode · Qwen/Qwen3.8-Max-0902
Provider
CommandCode*
Model
Qwen/Qwen3.8-Max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
  • pdf*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    effort*
    values
    • low*
    • medium*
    • xhigh*
Tool calling
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
API service · Kilo Gateway · qwen/qwen3.8-max-0902
Provider
Kilo Gateway*
Model
qwen/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    effort*
    values
    • minimal*
    • low*
    • medium*
    • high*
    • xhigh*
Tool calling
Yes*
Structured output
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
cache_write
2.5*
API service · OpenRouter · qwen/qwen3.8-max-0902
Provider
OpenRouter*
Model
qwen/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    effort*
    values
    • minimal*
    • low*
    • medium*
    • high*
    • xhigh*
Tool calling
Yes*
Structured output
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
API service · Eden AI · qwen/qwen3.8-max-0902
Provider
Eden AI*
Model
qwen/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
output
131,072*
Modalities
input
  • text*
  • image*
  • video*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    effort*
    values
    • none*
    • low*
    • medium*
    • xhigh*
Tool calling
Yes*
Structured output
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
cache_write
2.5*
API service · Vercel AI Gateway · alibaba/qwen3.8-max-0902
Provider
Vercel AI Gateway*
Model
alibaba/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
991,000*
output
128,000*
Modalities
input
  • text*
  • image*
  • pdf*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    toggle*
  • type
    effort*
    values
    • low*
    • medium*
    • xhigh*
  • type
    budget_tokens*
    min
    0*
    max
    262,144*
Tool calling
Yes*
Temperature control
Yes*
File attachments
Yes*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
cache_write
2.5*
API service · Standard Compute · qwen/qwen3.8-max-0902
Provider
Standard Compute*
Model
qwen/qwen3.8-max-0902*
Snapshot
2026-10-26*
Token limits
context
1,000,000*
input
850,000*
output
50,000*
Modalities
input
  • text*
  • image*
  • video*
  • pdf*
output
  • text*
Reasoning
Yes*
Reasoning controls
  • type
    effort*
    values
    • minimal*
    • low*
    • medium*
    • high*
    • xhigh*
Tool calling
Yes*
Temperature control
Yes*
File attachments
Yes*
Interleaved reasoning
field
reasoning_details*
Prices (USD per million tokens)
input
2*
output
6*
cache_read
0.25*
cache_write
2.5*
Reasoning
Yes*
Temperature control
Yes*
Tool calling
Yes*

* Provisional data