Inference
EN
InclusionAI

LLaDA2.2-mini

LLaDA2.2-mini is a reasoning model from InclusionAI in the LLaDA2.2 family. 9 benchmarks count toward its index, in 6 categories.

  • Open weights
  • Reasoning
  • Input: text
  • Released 2026-07
RankedEnough results for an index.
IndexAll categories, out of 100.24.3±6.9Median 51.0
CoverageIndex weight that has results.85%
SpeedOutput tokens per second.—Median 67/s
First tokenSeconds to the first output token.—Median 1.3 s
Input / 1MUS dollars per 1M input tokens.—Median $0.60
Output / 1MUS dollars per 1M output tokens.—Median $2.55
ContextMaximum tokens in one request.128KMedian 262K
EloLMArena rating and rank.—Median 1422

Run locally

16.26B parameters
16-bit
35.1 GB
8-bit
19.1 GB
4-bit
11.4 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy fits in 12 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.4 counted34.8189 / 218
CodingCode writing and repair.1 counted7.5—
ReasoningLogic problems and puzzles.1 counted21.9—
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.1 counted17.9—
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted12.3—
MathMath problems.1 counted10.5—

Results

9 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
τ²-Bench Tool-Agent-User EvaluationAgenticAbove fitFar above this model's index.57.5%88 / 149
Berkeley Function Calling Leaderboard v4Agentic47.7%19 / 26
Claw-EvalAgenticAbove fitFar above this model's index.57.2%33 / 42
GPQA Diamond · BenchLMKnowledge44.4%80 / 84
PinchBenchAgentic62.3%10 / 10
AIME 2026Math35.0%34 / 34
LongBench v2Reasoning35.0%25 / 25
LiveCodeBench v6Coding28.1%59 / 59
Instruction Following BenchmarkInstruction24.9%57 / 57

Sources

BenchLMCC BY-NC 4.0

History

5 releases
Index25.9 → 24.3
23 Oct 202612 Nov 2026

Questions

How good is LLaDA2.2-mini?
Its index is 24.3 out of 100. That is place 270 of 283 ranked models.
What is LLaDA2.2-mini best at?
Agentic: 34.8 out of 100, place 189 of 218.
What is the context window of LLaDA2.2-mini?
128K tokens.
Is LLaDA2.2-mini open source?
Yes. Its weights are published.
When was LLaDA2.2-mini released?
July 2026.

Badge

LLaDA2.2-mini on Inference360: place 270 of 283, index 24.3
HTML
<a href="https://inference360.ai/models/llada2-2-mini"><img src="https://inference360.ai/badges/llada2-2-mini.svg" alt="LLaDA2.2-mini on Inference360: place 270 of 283, index 24.3" height="20"></a>
Markdown
[![LLaDA2.2-mini on Inference360: place 270 of 283, index 24.3](https://inference360.ai/badges/llada2-2-mini.svg)](https://inference360.ai/models/llada2-2-mini)

Compare with

Ling 3.0 Flash50.9

More from InclusionAI

All models
LLaDA2.2-flashUnrankedLing 3.0 Flash50.9Ling 3.0 Flash FP8UnrankedLing 3.0 TinyUnrankedLing 3.0 Flash FinUnranked