Benchmark Almanac

Z.ai Code Bench

Claude Fable 5 39.5% leads since August 2026
In use

Zhipu in-house coding benchmark used for GLM-5.3. Reports pass rate and output tokens per task. Method unpublished.

The official launch posts publish exact accuracy and token values for selected effort levels. Additional effort points appear only in chart images.

Made by
Zhipu AI (Z.ai)
Released
August 2026
Kind
Fixing real repositories
Level
Expert
Scored
Checked automatically
Answers
Code
Tasks
Kept private
Strongest model tested
Claude Fable 5 #11 on the capability index
Last in a launch post
August 2026
Last model added
August 2026
Scores refreshed
November 12th, 2026
5
Models scored on its own leaderboard
1
Launch post lists it, from 1 lab

Claude Fable 5 leads with 39.5% among models released by August 2026

Best score per model from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best.

The newest model with a score here came out in August 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 39.5 Claude Fable 5
  2. 34.5 GLM-5.3
  3. 29.5 Claude Opus 4.8
  4. 29.0 GLM-5.3-Flash
  5. 23.4 GLM-5.2
  1. 1 Claude Fable 5 max 39.5%
  2. 2 GLM-5.3 max 34.5%
  3. 3 Claude Opus 4.8 max 29.5%
  4. 4 GLM-5.3-Flash max 29%
  5. 5 GLM-5.2 max 23.4%

Source: Official Z.ai Code Bench leaderboard

#ModelScoreSettingDateSource
1 Claude Fable 5 39.5% max, Claude Code 2.1.207 Jun 2026 Official leaderboard
2 GLM-5.3 34.5% max, Claude Code 2.1.207 Aug 2026 Official leaderboard
3 Claude Opus 4.8 29.5% max, Claude Code 2.1.207 May 2026 Official leaderboard
4 GLM-5.3-Flash 29% max, Claude Code 2.1.207 Aug 2026 Official leaderboard
5 GLM-5.2 23.4% max, Claude Code 2.1.207 Jun 2026 Official leaderboard

Labs report scores for 1 model in their cards. We plot every source in the charts below.

Agentic coding accuracy vs output tokens

Only exact numerical points stated in the article text are imported. Token figures are approximate as reported. The line joins the runs that no other run beats on both measures. We redrew this chart from the numbers on the official leaderboard.

25% 30% 35% 40% 60K80K100K120K Mean output tokens per task (tokens), lower is better Accuracy Claude Opus 4.8 (max): Accuracy 29.5%, mean output tokens per task 120K tokens GLM-5.2 (max): Accuracy 23.4%, mean output tokens per task 96K tokens GLM-5.3 (max): Accuracy 34.5%, mean output tokens per task 75K tokens GLM-5.3 (high): Accuracy 31.4%, mean output tokens per task 50K tokens GLM-5.3 (max)GLM-5.3 (high) 25% 30% 35% 40% 60K80K100K120K Mean output tokens per task (tokens), lower is better Accuracy Claude Opus 4.8 (max): Accuracy 29.5%, mean output tokens per task 120K tokens GLM-5.2 (max): Accuracy 23.4%, mean output tokens per task 96K tokens GLM-5.3 (max): Accuracy 34.5%, mean output tokens per task 75K tokens GLM-5.3 (high): Accuracy 31.4%, mean output tokens per task 50K tokens GLM-5.3 (max)GLM-5.3 (high)
ModelSettingAccuracyMean output tokens per task
GLM-5.3 max, Claude Code 2.1.207 34.5% 75,000 tokens
GLM-5.3 high, Claude Code 2.1.207 31.4% 50,000 tokens
Claude Opus 4.8 max, Claude Code 2.1.207 29.5% 120,000 tokens
GLM-5.2 max, Claude Code 2.1.207 23.4% 96,000 tokens

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Own leaderboard Not tested since Aug 2026
Claude Fable 5 39.5%

Source: Official Z.ai Code Bench leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Reported by the lab in its model card Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5 Claude Opus 4.8 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5 Claude Opus 4.8

Sources: Official Z.ai Code Bench leaderboard and labs’ model cards

One lab put it in a launch post in the last six months

Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.

Z.ai
DateLaunch postScoreAs the post writes it
Aug 26, 2026 GLM-5.3-Flash 29 Z.ai Code Bench v1.0, 29.0

One lab claim next to the leaderboard

Each score a lab reported for its own model, beside the same model's best score on the leaderboard. Gaps mostly come from different harnesses, effort levels and task subsets. 1 of them match, which usually means one side copied the other. Every benchmark.

Model Lab says Board says Gap
GLM-5.3-Flash 29% Z.ai Code Bench v1.0, 29.0 29% Leaderboard, max, Claude Code 2.1.207 Same

More in Coding

All 282 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471