Benchmark Almanac

LIBRA (Russian Long Context)

Qwen3-30B-A3B-Instruct Jul 2025 71.3 leads since July 2025
Not in model cards

Models read long Russian texts and answer questions or generate responses. The suite groups tasks by difficulty and varies input length.

Made by
Igor Churin and colleagues
Released
August 2024
Kind
Reasoning over long documents
Level
University
Scored
Checked automatically
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
Russian
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Qwen3.5-35B-A3B #103 on the capability index
Last model added
February 2026
Scores refreshed
November 20th, 2026
13
Entries on its own leaderboard, 9 models we track

Qwen3-30B-A3B-Instruct leads with 71.3 among models released by February 2026 Qwen3-30B-A3B-Instruct leads with 71.3 among models released by February 2026

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. This view shows only models we track, each at its best.

The newest model with a score here came out in February 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 Qwen3-30B-A3B-Instruct Jul 2025 71.3
  2. 2 Qwen3.5-122B-A10B 65.0
  3. 3 meta-llama_Llama-3.1-8B-Instruct 64.7
  4. 4 Qwen3.5-35B-A3B 61.0
  5. 5 Qwen3.5-9B 51.0
  6. 6 01-ai_Yi-9B-200K 49.2
  7. 7 Phi-3.5-mini 48.8
  8. 8 01-ai_Yi-34B-200K 47.6
  9. 9 gradientai_Llama-3-8B-Instruct-Gradient-1048k 45.6
  10. 10 Qwen3.5-4B 43.0
  11. 11 Qwen3.5-2B 34.0
  12. 12 LFM2.5-1.2B-Instruct 28.8
  13. 13 Qwen3.5-0.8B 27.0

Source: Official LIBRA (Russian Long Context) leaderboard

#EntryScoreSettingDateSource
1 Qwen3-30B-A3B-Instruct Jul 2025 71.3 Jul 2025 Official leaderboard
2 Qwen3.5-122B-A10B 65.0 Feb 2026 Official leaderboard
3 meta-llama_Llama-3.1-8B-Instruct 64.7 Official leaderboard
4 Qwen3.5-35B-A3B 61.0 Feb 2026 Official leaderboard
5 Qwen3.5-9B 51.0 Feb 2026 Official leaderboard
6 01-ai_Yi-9B-200K 49.2 Official leaderboard
7 Phi-3.5-mini 48.8 Aug 2024 Official leaderboard
8 01-ai_Yi-34B-200K 47.6 Official leaderboard
9 gradientai_Llama-3-8B-Instruct-Gradient-1048k 45.6 Official leaderboard
10 Qwen3.5-4B 43.0 Feb 2026 Official leaderboard
11 Qwen3.5-2B 34.0 Feb 2026 Official leaderboard
12 LFM2.5-1.2B-Instruct 28.8 Jan 2026 Official leaderboard
13 Qwen3.5-0.8B 27.0 Feb 2026 Official leaderboard
#ModelScoreSettingDateSource
1 Qwen3-30B-A3B-Instruct Jul 2025 71.3 Jul 2025 Official leaderboard
2 Qwen3.5-122B-A10B 65.0 Feb 2026 Official leaderboard
3 Qwen3.5-35B-A3B 61.0 Feb 2026 Official leaderboard
4 Qwen3.5-9B 51.0 Feb 2026 Official leaderboard
5 Phi-3.5-mini 48.8 Aug 2024 Official leaderboard
6 Qwen3.5-4B 43.0 Feb 2026 Official leaderboard
7 Qwen3.5-2B 34.0 Feb 2026 Official leaderboard
8 LFM2.5-1.2B-Instruct 28.8 Jan 2026 Official leaderboard
9 Qwen3.5-0.8B 27.0 Feb 2026 Official leaderboard

Rankings by part

We read 24 parts from the benchmark’s own leaderboard in 2 groups: task and context length. Each ranks models on one part of the test. Qwen3-30B-A3B-Instruct leads 23 of them, and Phi-3.5-mini leads the other one. Arrows show how far a model moved from its overall rank among the models we track.

Passkey

9 models we track have a score here.

  1. 1 Qwen3-30B-A3B-Instruct Jul 2025 100.0
  2. 2 Phi-3.5-mini ↑3 99.8
  3. 3 Qwen3.5-35B-A3B 84.7
  4. 4 Qwen3.5-9B 80.6
  5. 5 Qwen3.5-122B-A10B ↓3 80.2
  6. 6 Qwen3.5-4B 75.9
  7. 7 Qwen3.5-2B 72.6
  8. 8 Qwen3.5-0.8B ↑1 69.0
  9. 9 LFM2.5-1.2B-Instruct ↓1 66.7

2 models have held number one on the official leaderboard since its release in August 2024

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

Microsoft Qwen
Own leaderboard Not tested since Feb 2026

Source: Official LIBRA (Russian Long Context) 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
20.0 40.0 60.0 80.0 2025 2026 20.0 40.0 60.0 80.0 2025 2026

Source: Official LIBRA (Russian Long Context) 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 Long context

All 115 in Long context
  • MRCR v2 In use Multi-round coreference: retrieve the n-th of several similar requests hidden in a long synthetic conversation, up to 1M tokens. Muse Spark 1.3 98.5%
  • LongBench v2 In use 503 multiple-choice questions over 8K-2M-word contexts (QA, dialogue, code repos, structured data). Accuracy. Inkling 2 Small 70.6%
  • GraphWalks In use A graph is given as an edge list in long context. The model answers BFS or parent queries, scored by F1. Claude Fable 5 99.96%
  • OpenAI MRCR v2 In use Models read a long synthetic chat that hides 2, 4 or 8 identical writing requests and must reproduce the response to the i-th one. Scored by mean string match ratio per context bin. Gemini 3.7 Flash 89%
  • CL-bench In use Model must learn new knowledge from a supplied long context (rules, procedures, data) and solve tasks. Strict LLM-verified rubrics. GPT-5.4 27.9%
  • AA-LCR 1.1 In use Models answer 100 questions that need reasoning across sets of real documents of about 100K tokens. An LLM equality checker with a fixed system prompt compares each answer with the key. Step 5 Preview 88.3%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 9 months ago, in February 2026, for Qwen3.5-0.8B. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (13 entries).
  • Pass Models scored 9 models we track, plus 4 other entries on its leaderboard.
  • Warning Newest score The newest came 9 months ago, in February 2026, for Qwen3.5-0.8B.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Qwen3.5-35B-A3B (#103).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on huggingface.co: 9 rows for models we track and 4 other entries 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