Intellect 3 — Retired model
INTELLECT-3 is a 106B-parameter Mixture-of-Experts model (12B active) post-trained from GLM-4.5-Air-Base using supervised fine-tuning (SFT) followed by large-scale reinforcement learning (RL). It offers state-of-the-art performance for its size across math,...
Technical specifications
Independent benchmarks
Scores below come from Artificial Analysis, an independent third party. BazaarLink does not participate in the testing.
Intelligence 16
Coding
Math 88
MMLU 82
GPQA 76
This modelCategory leader
Overall intelligence
#279out of 573 models
Aggregated across academic benchmarks
Math reasoning
#37out of 269 models
AIME / MATH-500 / SciCode
Time to first token
🥇
#1out of 586 models
AA 全球 p50：0.00 秒
Intelligence
16
Mid-tierThis model16
Category leader61
Median15
Math
88
Top 14%This model88
Category leader99
Median53
MMLU Pro
82%
—This model82%
Category leader90%
Median75%
GPQA
76%
—This model76%
Category leader94%
Median69%
LiveCodeBench
78%
—This model78%
Category leader42%
Median42%
HLE
12%
—This model12%
Category leader53%
Median7%
SciCode
39%
—This model39%
Category leader60%
Median33%
IFBench
34%
—This model34%
Category leader83%
Median44%
τ²-Bench Telecom
27%
—This model27%
Category leader99%
Median46%
AA-LCR
32%
—This model32%
Category leader76%
Median40%
Terminal-Bench Hard
9%
—This model9%
Category leader66%
Median14%
Top 5 — CodingCoding Index leaderboard
1GPT-5.6 Sol (xhigh)
openai/gpt-5-6-sol-xhigh78$5 / $304GPT-5.6 Sol (high)
openai/gpt-5-6-sol-high77$5 / $305Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)
anthropic/claude-opus-5-xhigh77$5 / $25AASource: Artificial Analysis · Independent third-party evaluation; not influenced by BazaarLinkView full evaluation on AA →
Frequently Asked Questions
Can I still use the retired Intellect 3 model?
No. Intellect 3 has been retired and is retained here only as a historical reference. See the active alternatives linked on this page.
Relay integrity checks
No integrity-check data is available for this model yet.
BazaarLink Probe verifies endpoints that claim to serve this model and flags behavior that differs from the reference baseline.
View the full integrity report →Related models and links
Related models and links