Skip to main content
Radar

Keep up with the models you depend on. Follow price changes, retirements, and API updates.Follow the models you depend on.

Follow model changes

LLaDA2.2-mini

Released Jul 16, 2026 see all recent releases

Decision reading
LLaDA2.2-mini is tracked, but not publicly ranked yet. The profile exposes 9 sourced benchmark rows and leaves unsupported fields blank until a published record exists.

Data as of September 10, 2026 · How the score is built

Strongest published evidence

Published rows are visible, but no category has enough eligible evidence for a comparative rank.

Validate before choosing

9 published rows leave some tracked benchmark slots empty. No comparable first-party API token rate is published.

Decision snapshot

Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.

Capability

Unranked

field median 56.2

Not eligible for a public rank

Price

Self-hosted; infrastructure cost varies

input median $1

No comparable first-party hosted token rate

Speed

Not measured

field median 86.5 tok/s

Time to first token not measured

Context

128Ktokens

field median 256,000

Maximum output length is tracked separately

How much of this is verified

Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.

  1. Agentic4/4 verified
  2. Coding1/1 verified
  3. Reasoning1/1 verified
  4. Knowledge1/1 verified
  5. Math1/1 verified
  6. MultilingualNot measured
  7. MultimodalNot measured
  8. Inst. Following1/1 verified
Verified sourceProvisionalNot measured

Spec sheet

Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.

API model ID
inclusionAI/LLaDA2.2-miniInclusionAI LLaDA2.2-mini model card
Maximum output
Not sourced yet
Knowledge cutoff
Not sourced yet
Input modalities
Not sourced yet
Output modalities
Not sourced yet
Parameters
Not sourced yet
Availability
InclusionAI publishes the 16B lightweight agentic MoE diffusion model under Apache 2.0 on Hugging Face with a 128K context window and block routing.
Cloud regions
Not tracked yet
Lifecycle
Current
API capabilities
Tool calling, structured outputs, and batch support are not tracked yet
Prompt caching
Not documented in the pricing recordInclusionAI LLaDA2.2-mini model card
Self-host
Open weights available; hardware estimate not sourced
Rate limits
Not tracked yet

Deployment options

Self-host and provider-specific paths stay separate from benchmark evidence so operating constraints are visible before a score becomes the whole decision.

Published weights are available, but BenchLM does not yet have a sourced parameter and VRAM profile for this exact model. Hardware cost estimates stay unavailable until that sizing record is complete.

Estimate VRAM from known parameters

Category score record

Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.

Category scores, ranks, weighting, benchmark coverage, and evidence status
CategoryScore
AgenticWeight 22%4 benchmarksVerifiedScore pending
CodingWeight 20%1 benchmarkVerifiedScore pending
ReasoningRank Not rankedWeight 17%1 benchmarkVerified23.8
KnowledgeWeight 12%1 benchmarkVerifiedScore pending
MathRank Not rankedWeight 5%1 benchmarkVerified17.3
MultilingualWeight 7%0 benchmarksNot measuredNot measured
MultimodalWeight 12%0 benchmarksNot measuredNot measured
Inst. FollowingRank Not rankedWeight 5%1 benchmarkVerified9.7

Benchmark ledger

Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.

Coding1 row
Coding benchmark values, best verified comparison, weight, and source status
LiveCodeBench v6Score28.1%Versus best verified row

Best verified: Sakana Fugu-Ultra · 93.2%

Gap65.1 behindWeightDisplay only
Agentic4 rows
Agentic benchmark values, best verified comparison, weight, and source status
BFCL v4Berkeley Function Calling Leaderboard v4Score47.7%Versus best verified row

Best verified: BTL-3 · 88.5%

Gap40.8 behindWeightDisplay only
τ²-bench resultsτ²-Bench Tool-Agent-User EvaluationScore57.5%Versus best verified row

Best verified: GPT-5.4 · 98.9%

Gap41.4 behindWeightDisplay only
Claw-EvalScore57.2%Versus best verified row

Best verified: Ornith-1.5-397B · 81.4%

Gap24.2 behindWeightDisplay only
PinchBenchScore62.3%Versus best verified row

Best verified: Pokee-Isaac 28B · 95.7%

Gap33.3 behindWeightDisplay only
Reasoning1 row
Reasoning benchmark values, best verified comparison, weight, and source status
LongBench v2Score35.0%Versus best verified row

Best verified: Qwen3.8 Max · 66.3%

Gap31.3 behindWeightWeighted 25%
Knowledge1 row
Knowledge benchmark values, best verified comparison, weight, and source status
GPQA-DGPQA DiamondScore44.4%Versus best verified row

Best verified: GPT-6 Astra · 96.0%

Gap51.6 behindWeightDisplay only
Math1 row
Math benchmark values, best verified comparison, weight, and source status
AIME26AIME 2026Score35.0%Versus best verified row

Best verified: GLM-5.2 · 99.2%

Gap64.2 behindWeightWeighted 25%
Inst. Following1 row
Inst. Following benchmark values, best verified comparison, weight, and source status
IFBenchInstruction Following BenchmarkScore24.9%Versus best verified row

Best verified: MAI-Thinking-1 · 85%

Gap60.1 behindWeightWeighted 70%

Lineage

The sequence follows explicit supersedes links. A successor's displayed score stays at least 0.1 points above its predecessor; raw benchmark rows do not move. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.

  1. Jul 16, 2026 · you are here

    LLaDA2.2-mini

    Not publicly ranked · Price not listed

Mini

How to read this profile

The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.

We track LLaDA2.2-mini, but the public leaderboard excludes this profile until enough non-generated benchmark coverage is available. Only published rows appear above.

LLaDA2.2-mini is a open weight model with a 128K context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.

InclusionAI publishes the 16B lightweight agentic MoE diffusion model under Apache 2.0 on Hugging Face with a 128K context window and block routing.

Official exact-value snapshot from the LLaDA2.2-mini Hugging Face model card (September 5, 2026). BenchLM maps BFCL v4, AIME 2026, LiveCodeBench v6, IFBench, GPQA-Diamond, LongBench v2, τ²-Bench, Claw-Eval, and PinchBench onto existing lanes; BFCL v3, OlympiadBench, MultiPL-E, Multi-IF, and KOR-Bench have no exact compatible lane. Provider-run rows keep the row unranked.

LLaDA2.2-mini sits in the LLaDA2.2 family with LLaDA2.2-flash. 9 of 434 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.

Radar

LLaDA2.2-mini release history

Full release history

Radar confirmed these at the source. Use LLaDA2.2-mini in your work? Explore Radar to follow supported changes and choose your alerts.

Frequently asked questions

How does LLaDA2.2-mini perform overall in AI benchmarks?

LLaDA2.2-mini has 9 source-displayable benchmark rows, but it does not qualify for a public overall rank. The available rows remain visible by category without being converted into a site-wide score. Missing evidence stays blank instead of being estimated from an earlier model.

Is LLaDA2.2-mini good for knowledge and understanding?

LLaDA2.2-mini has source-displayable benchmark coverage for knowledge and understanding, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

Is LLaDA2.2-mini good for coding and programming?

LLaDA2.2-mini has source-displayable benchmark coverage for coding and programming, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

Is LLaDA2.2-mini good for mathematics?

LLaDA2.2-mini has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

Is LLaDA2.2-mini good for reasoning and logic?

LLaDA2.2-mini has source-displayable benchmark coverage for reasoning and logic, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

Is LLaDA2.2-mini good for agentic tool use and computer tasks?

LLaDA2.2-mini has source-displayable benchmark coverage for agentic tool use and computer tasks, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

Is LLaDA2.2-mini good for instruction following?

LLaDA2.2-mini has source-displayable benchmark coverage for instruction following, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.

Is LLaDA2.2-mini open source?

LLaDA2.2-mini is an open-weight model from InclusionAI. Its weights can be downloaded for local or hosted deployment, subject to the published license. Open weight does not automatically mean open source: training data and training code may remain private, and commercial restrictions can still apply.

Which sibling models are related to LLaDA2.2-mini?

LLaDA2.2-mini belongs to the LLaDA2.2 family. Related tracked variants include LLaDA2.2-flash. A sibling link indicates shared lineage or a documented configuration relationship; it does not mean the variants have identical pricing, context limits, benchmark evidence, or deployment behavior. Compare before switching.

Does LLaDA2.2-mini have full benchmark coverage on BenchLM?

No. LLaDA2.2-mini currently has 9 source-displayable rows across 434 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.

What is the context window size of LLaDA2.2-mini?

LLaDA2.2-mini has a documented context window of 128K. That figure is the maximum combined prompt and retained-conversation space reported for this exact model; it is not the maximum output length. The profile keeps output limits separate because providers often publish those limits independently.

Compare LLaDA2.2-mini with every tracked model482 comparisons

Last updated September 10, 2026. Runtime fields remain blank until a sourced snapshot exists.

Watch LLaDA2.2-mini in the weekly brief

Get one weekly email when material rank, price, availability, or benchmark evidence changes are worth revisiting.

Read a sample issue

Join 2,000+ readers.