[whatmodel.sh](../../README.md) / GigaChat 3.5 Ultra Reasoning (432B-A28B) # GigaChat 3.5 Ultra Reasoning (432B-A28B) Sber · released 2026-09 · 438.09B params (28B active) · mit **score 65.8** ±1.3 · #131 of 732 · from 10 benchmark results across 1 source · † no independently run results verified Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at _{● marks GigaChat 3.5 Ultra Reasoning (432B-A28B) among every model on that task or test, worst to best.}

task         vs the others          rank  counts  tests
agent tasks  ▁▁▃▆█▄●▅▄▄▂▁    #134 of 313     27%      3
coding       ▁▁▁▃▅▇█▇●▄▃▁   #122 of 469*     23%      2
math         ▁▁▃▅█▆▄●▄▂▁▁     #90 of 431     41%      4

agent tasks · #134 of 313 · 27% of its score

Beats [o3](o3.md), [Gemini 2.5 Pro](gemini-2-5-pro.md), [Gemini 3.1 Flash-Lite](gemini-3-1-flash-lite.md) here, though they rank higher overall. Trails [SmolLM3 3B](smollm3-3b.md), [Jamba 1.5 Large](jamba-1-5-large.md), [Devstral 2 (2512)](devstral-2.md) here, though they rank lower overall.

test                        result  vs the others in its run   difficulty           counts  why
SWE-bench Verified (maker)   64.7%  ▁▁▁▁▁▁▂▄●▆▁▁ beats 68/138  hard · avg 24%          14%  sharp · at its level · small run
Terminal-Bench 2.0 (maker)   30.3%  ▂▂▅●████▆▂▃▃ beats 5/36    very hard · avg 10%     10%  at its level · small run
τ³-bench (maker)             47.8%  ▁▁▂▁·▁▁·●▁▂█ beats 9/25    hard · avg 34%         3.0%  at its level · small run

coding · #122 of 469 · 23% of its score

Beats [Qwen3.5 Plus](qwen3-5-plus.md), [Qwen3-Coder-Next (80B-A3B)](qwen3-coder-next.md), [Gemini 2.5 Pro](gemini-2-5-pro.md) here, though they rank higher overall. Trails [Llama 3.1 Nemotron Ultra 253B v1](llama-3-1-nemotron-ultra-253b.md), [Solar Pro 4](solar-pro-4.md), [Laguna XS 2.1 (33B-A3B)](laguna-xs-2-1.md) here, though they rank lower overall.

test                                result  vs the others in its run   difficulty        counts  why
SWE-bench Verified (maker)           64.7%  ▁▁▁▁▁▁▂▄●▆▁▁ beats 68/138  hard · avg 24%       14%  sharp · at its level · small run
LiveCodeBench v6 (window unspecif…   85.4%  ▂▃▄▄▅██▆▄●▄▆ beats 53/67   medium · avg 47%     10%  small run

math · #90 of 431 · 41% of its score

Beats [Claude Sonnet 4.6](claude-sonnet-4-6.md), [Claude Opus 4.5](claude-opus-4-5.md), [MiniMax-M3](minimax-m3.md) here, though they rank higher overall. Trails [Gemini 2.5 Flash](gemini-2-5-flash.md), [Granite 4.2 30B](granite-4-2-30b.md), [Qwen3-Next 80B-A3B Instruct](qwen3-next-80b-a3b-instruct.md) here, though they rank lower overall.

test                                result  vs the others in its run    difficulty        counts  why
AIME 2025 (maker)                    89.0%  ▁·▁▁▁▂▅▆█●▅▁ beats 110/133  medium · avg 48%     11%  sharp · small run
AIME 2026 (maker)                    92.0%  ▁▁▁▂▂▂▆●▆▂·▁ beats 29/52    medium · avg 53%     11%  sharp · small run
HMMT 2025 (edition unspecified) (…   83.1%  ▃█▇▅▇●▁·▁··▁ beats 26/32    hard · avg 31%       10%  sharp · small run
IMO-AnswerBench (maker)              73.0%  ▁▁·▁▁▁●▆█▄▂▃ beats 12/44    hard · avg 31%      9.2%  at its level · small run

other tests (2)

test                                result  vs the others in its run    difficulty        counts  why
GPQA Diamond (maker)                 82.3%  ▁▁▁▁▂▄▅▅█●█▄ beats 218/322  hard · avg 44%       17%  sharp · small run
IFBench (aggregation unspecified)…   77.0%  ▂▆▂▄▄▄▂▄▄▆●▇ beats 74/86    medium · avg 40%    4.7%  at its level · small run

## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md)

source                                    results  share of the score
ai-sage/GigaChat3.5-432B-A28B-Reasoning…       10   99% ▇▇▇▇▇▇▇▇▇▇▇▇
prior: pull toward the middle                        1% ▇
weighted average                                   = 65.8

_{Licences: ai-sage/GigaChat3.5-432B-A28B-Reasoning model card ([MIT](https://huggingface.co/ai-sage/GigaChat3.5-432B-A28B-Reasoning/blob/02f55379ac0b20abd9378791bffd6254725e6f5c/README.md))} ## How to run it ### Other ways to run it 1. Run it as published (fp8, needs ~438 GB, NVIDIA only, from [huggingface.co/ai-sage](https://huggingface.co/ai-sage/GigaChat3.5-432B-A28B-Reasoning/tree/02f55379ac0b20abd9378791bffd6254725e6f5c)) ```sh vllm serve ai-sage/GigaChat3.5-432B-A28B-Reasoning --tensor-parallel-size 8 ``` Model repository and documentation: [huggingface.co/ai-sage](https://huggingface.co/ai-sage/GigaChat3.5-432B-A28B-Reasoning) [interactive version](https://whatmodel.sh/model/gigachat-3-5-ultra-reasoning) · [JSON](https://whatmodel.sh/api?format=json&all=1)