Benchmark Almanac

Itzi Roleplay Benchmark: Local Robustness

MiMo-V2.5 100.0 leads since September 2026
Not in model cards

Models play eight scripted episodes. Judges check whether each model remembers instructions, respects player decisions, and keeps characters and locations consistent.

The maker reports local robustness for 36 models. It counts missing episode criteria as failures. External Arena scores supply the other 75% of overall quality.

Made by
Itzi
Released
September 2026
Kind
Creative writing
Level
Everyday
Scored
Graded by a model
Answers
Open-ended
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Multi-turn
Refresh
One fixed set
Run by
A host
Strongest model tested
GPT-6 Astra #4 on the capability index
Last model added
September 2026
Scores refreshed
November 20th, 2026
36
Models scored on its own leaderboard

6 models share the top score of 100.0 among models released by September 2026

Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

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

  1. 1 MiMo-V2.5 100.0
  2. 2 MiMo-V2.5-Pro 100.0
  3. 3 Claude Opus 4.8 100.0
  4. 4 Qwen 3.8 Max 100.0
  5. 5 GPT-6 Astra 100.0
  6. 6 Claude Opus 5.5 100.0
  7. 7 GPT-5.5 94.9
  8. 8 GPT-5.6 Sol 94.9
  9. 9 Gemini 3.6 Flash thinking 94.9
  10. 10 GLM-5.3 94.9
  11. 11 GLM-5.3-Flash 94.9
  12. 12 Gemini 3.8 Flash thinking 94.9
  13. 13 MiMo-V2.6 Pro 94.9
  14. 14 Claude Opus 4.7 92.3
  15. 15 MiniMax-M3 92.3
  16. 16 GPT-5.6 Terra 92.3
  17. 17 Claude Opus 5 92.3
  18. 18 Claude Fable 5.1 92.3
  19. 19 Claude Sonnet 4.6 87.2
  20. 20 Gemini 3.1 Pro 87.2
  21. 21 GLM-5.1 87.2
  22. 22 Claude Sonnet 5 87.2
  23. 23 Qwen 3.8 27B 84.6
  24. 24 Claude Fable 5 79.5
  25. 25 GPT-5.6 Luna 79.5
  26. 26 Kimi K3 79.5
  27. 27 Gemini 3.7 Flash thinking 77.8
  28. 28 DeepSeek-V4-Flash 76.9
  29. 29 DeepSeek-V4-Pro 76.9
  30. 30 MiMo-V2.6 Flash 76.9
  31. 31 Claude Opus 4.6 71.8
  32. 32 Gemma 4 31B IT 71.8
  33. 33 Gemma 4 26B A4B 66.7
  34. 34 GLM-5.2 64.1
  35. 35 Grok 4.6 25.6
  36. 36 Grok 4.5 12.8

Source: Official Itzi Roleplay Benchmark: Local Robustness leaderboard

#ModelScoreSettingDateSource
1 MiMo-V2.5 100.0 Apr 2026 Official leaderboard
2 MiMo-V2.5-Pro 100.0 Apr 2026 Official leaderboard
3 Claude Opus 4.8 100.0 May 2026 Official leaderboard
4 Qwen 3.8 Max 100.0 Aug 2026 Official leaderboard
5 GPT-6 Astra 100.0 Sep 2026 Official leaderboard
6 Claude Opus 5.5 100.0 Sep 2026 Official leaderboard
7 GPT-5.5 94.9 Apr 2026 Official leaderboard
8 GPT-5.6 Sol 94.9 Jul 2026 Official leaderboard
9 Gemini 3.6 Flash 94.9 thinking Jul 2026 Official leaderboard
10 GLM-5.3 94.9 Aug 2026 Official leaderboard
11 GLM-5.3-Flash 94.9 Aug 2026 Official leaderboard
12 Gemini 3.8 Flash 94.9 thinking Sep 2026 Official leaderboard
13 MiMo-V2.6 Pro 94.9 Sep 2026 Official leaderboard
14 Claude Opus 4.7 92.3 Apr 2026 Official leaderboard
15 MiniMax-M3 92.3 Jun 2026 Official leaderboard
16 GPT-5.6 Terra 92.3 Jul 2026 Official leaderboard
17 Claude Opus 5 92.3 Jul 2026 Official leaderboard
18 Claude Fable 5.1 92.3 Sep 2026 Official leaderboard
19 Claude Sonnet 4.6 87.2 Feb 2026 Official leaderboard
20 Gemini 3.1 Pro 87.2 Feb 2026 Official leaderboard
21 GLM-5.1 87.2 Apr 2026 Official leaderboard
22 Claude Sonnet 5 87.2 Jun 2026 Official leaderboard
23 Qwen 3.8 27B 84.6 Aug 2026 Official leaderboard
24 Claude Fable 5 79.5 Jun 2026 Official leaderboard
25 GPT-5.6 Luna 79.5 Jul 2026 Official leaderboard
26 Kimi K3 79.5 Jul 2026 Official leaderboard
27 Gemini 3.7 Flash 77.8 thinking Aug 2026 Official leaderboard
28 DeepSeek-V4-Flash 76.9 Apr 2026 Official leaderboard
29 DeepSeek-V4-Pro 76.9 Apr 2026 Official leaderboard
30 MiMo-V2.6 Flash 76.9 Sep 2026 Official leaderboard
31 Claude Opus 4.6 71.8 Feb 2026 Official leaderboard
32 Gemma 4 31B IT 71.8 Apr 2026 Official leaderboard
33 Gemma 4 26B A4B 66.7 Apr 2026 Official leaderboard
34 GLM-5.2 64.1 Jun 2026 Official leaderboard
35 Grok 4.6 25.6 Aug 2026 Official leaderboard
36 Grok 4.5 12.8 Jul 2026 Official leaderboard

The best score, 100.0, costs $0.005 for the full run

Runs on the cost frontier score higher than any run that's cheaper. The costs are the leaderboard's own estimates, so compare them only within this benchmark.

Xiaomi Corp Cost frontier One model at each effort level
0.00 20.0 40.0 60.0 80.0 100.0 $0.001 $0.01 $0.10 $1.00 $10 Cost of the full run, log scale MiMo-V2.5 MiMo-V2.6 Flash 0.00 20.0 40.0 60.0 80.0 100.0 $0.001 $0.01 $0.10 $1.00 $10 Cost of the full run, log scale MiMo-V2.5 MiMo-V2.6 Flash

Source: Official Itzi Roleplay Benchmark: Local Robustness leaderboard

ModelEffortHarnessScoreCost of the full runOutput tokens
MiMo-V2.5 100.0 $0.005
MiMo-V2.5-Pro 100.0 $0.01
Claude Opus 4.8 100.0 $0.94
Qwen 3.8 Max 100.0 $0.46
GPT-6 Astra 100.0 $1.40
Claude Opus 5.5 100.0 $0.88
GPT-5.5 94.9 $0.60
GPT-5.6 Sol 94.9 $1.02
Gemini 3.6 Flash thinking 94.9 $0.14
GLM-5.3 94.9 $0.29
GLM-5.3-Flash 94.9 $0.04
Gemini 3.8 Flash thinking 94.9 $0.08
MiMo-V2.6 Pro 94.9 $0.02
Claude Opus 4.7 92.3 $1.18
MiniMax-M3 92.3 $0.02
GPT-5.6 Terra 92.3 $0.38
Claude Opus 5 92.3 $0.75
Claude Fable 5.1 92.3 $1.35
Claude Sonnet 4.6 87.2 $0.17
Gemini 3.1 Pro 87.2 $0.32
GLM-5.1 87.2 $0.07
Claude Sonnet 5 87.2 $0.85
Qwen 3.8 27B 84.6 $0.16
Claude Fable 5 79.5 $2.37
GPT-5.6 Luna 79.5 $0.04
Kimi K3 79.5 $0.20
Gemini 3.7 Flash thinking 77.8 $0.05
DeepSeek-V4-Flash 76.9 $0.01
DeepSeek-V4-Pro 76.9 $0.03
MiMo-V2.6 Flash 76.9 $0.004
Claude Opus 4.6 71.8 $0.37
Gemma 4 31B IT 71.8 $0.006
Gemma 4 26B A4B 66.7 $0.005
GLM-5.2 64.1 $0.07
Grok 4.6 25.6 $0.06
Grok 4.5 12.8 $0.03

Number one over time

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

Xiaomi Corp
Own leaderboard Not tested since Sep 2026
MiMo-V2.5 100.0

Source: Official Itzi Roleplay Benchmark: Local Robustness leaderboard

Scores over time

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

Plot with other benchmarks
On the benchmark’s own leaderboard
0.00 20.0 40.0 60.0 80.0 100.0 2026 0.00 20.0 40.0 60.0 80.0 100.0 2026

Source: Official Itzi Roleplay Benchmark: Local Robustness leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Writing and taste

All 65 in Writing and taste
  • Design Bench (Moonshot) In use Models make graphic designs from direct prompts or few-shot reference images. An automated grader rates the results as Elo scores. Kimi K3.1 Pro 957
  • WritingBench Fading Models complete 1,000 real-world writing requests across 6 domains and 100 subdomains. A judge model scores each reply on five query-specific criteria. GPT-5 83.9
  • Creative Writing (MiniMax in-house) Only in system cards Models write creative pieces for real Hailuo AI user requests. Automatic assessments score the writing. MiniMax-Text-01 81.3%
  • DeepSeek Chinese functional writing evaluation Only in system cards Models answer 3,170 everyday Chinese writing prompts such as reports, emails and speeches. Pairwise judging against a baseline model gives a win rate. DeepSeek-V4-Pro 62.6%
  • Slide generation human evaluation (Z.ai) Only in system cards People compare HTML slide decks from two models on content, layout and visual quality. Metric is overall win rate. GLM-5 67.5%
  • Task-dependent output homogenization evaluation Only in system cards Models answer prompts several times. A judge scores functional diversity where it is wanted and consistency where one answer is right. Muse Spark 0.21

Data audit

9 of 10 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (36 entries).
  • Pass Models scored 36 models we track.
  • Pass Newest score The newest is from September 2026, for Claude Opus 5.5.
  • Pass Frontier models 3 of the top 10 have a score, led by GPT-6 Astra (#4).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on itzi.app: 36 rows for models we track at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check