Granite 4.0 Micro vs Llama 3.2 1B Instruct
Granite 4.0 Micro leads Llama 3.2 1B Instruct on the Artificial Analysis Intelligence Index, 5.1 to 4.8, and Granite 4.0 Micro costs 1.8x less per output token ($0.112 vs $0.201 per million). Updated October 22, 2026; 13 benchmarks have a result for both.
Granite 4.0 Micro or Llama 3.2 1B Instruct: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Granite 4.0 Micro for output tokens 1.8x cheaper ($0.112 vs $0.201 per million) and a larger context window (131K against 60K tokens).
- Shared results
- 13
- Granite 4.0 Micro only
- 0
- Llama 3.2 1B Instruct only
- 25
- Like-for-like categories
- 7 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Granite 4.0 Micro
Granite 4.0 Micro leads, 5.1 against 4.8.
Same board, both measured
Long documents
Listed context window
Granite 4.0 Micro
Granite 4.0 Micro takes 131K tokens against 60K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Granite 4.0 Micro
Granite 4.0 Micro $0.0001, Llama 3.2 1B Instruct $0.0001 per turn.
List prices
At a glance
- Coding index
- Granite 4.0 Micro not benchmarked, Llama 3.2 1B Instruct not benchmarked.
- Widest category
- Knowledge, 24.7 points apart on average over 1 shared benchmark.
- Widest single gap
- MMLU Pro (Artificial Analysis run): Granite 4.0 Micro 44.7% against Llama 3.2 1B Instruct 20.0%, 24.7 points apart on a 0 to 100 scale.
- Output price
- Granite 4.0 Micro $0.112 against Llama 3.2 1B Instruct $0.201 per million output tokens, 1.8x apart.
- Context window
- Granite 4.0 Micro takes 131K tokens, Llama 3.2 1B Instruct 60K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Granite 4.0 Micro | Bars | Llama 3.2 1B Instruct | Ties |
|---|---|---|---|---|
| Coding | 2 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 2 | 1 | ||
| Math | 3 | 0 | ||
| Knowledge | 1 | 0 | ||
| General | 2 | 0 | ||
| All shared boards | 12 | 1 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index20
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| LiveCodeBench (Artificial Analysis run) | 18.0% | 1.9% | Granite 4.0 Micro |
| Terminal-Bench Hard (Artificial Analysis run) | 1.5% | 0.0% | Granite 4.0 Micro |
| LMArena Text: coding | Not measured | 1070 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: software and IT | Not measured | 1056 | Only Llama 3.2 1B Instruct measured |
| Neither model has a result on the other 39 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | 5.1 | 4.8 | Granite 4.0 Micro |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 12.6% | 0.0% | Granite 4.0 Micro |
| BALROG | Not measured | 6.6% | Only Llama 3.2 1B Instruct measured |
| Berkeley Function Calling Leaderboard (BFCL) | Not measured | 10.8% | Only Llama 3.2 1B Instruct measured |
| Neither model has a result on the other 39 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index21
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 33.6% | 19.6% | Granite 4.0 Micro |
| Humanity's Last Exam (Artificial Analysis run) | 5.0% | 5.5% | Llama 3.2 1B Instruct |
| GPQA Diamond (Epoch AI run) | 28.3% | 23.9% | Granite 4.0 Micro |
| LMArena Text: hard prompts | Not measured | 1044 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: expert | Not measured | 1007 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: hard prompts in English | Not measured | 1087 | Only Llama 3.2 1B Instruct measured |
| Neither model has a result on the other 17 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| AA Math Index | 6.0 | 0.0 | Granite 4.0 Micro |
| AIME 2025 (Artificial Analysis run) | 6.0% | 0.0% | Granite 4.0 Micro |
| OTIS Mock AIME | 2.8% | 0.6% | Granite 4.0 Micro |
| LMArena Text: math | Not measured | 1086 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: mathematical work | Not measured | 1098 | Only Llama 3.2 1B Instruct measured |
| Neither model has a result on the other 9 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 44.7% | 20.0% | Granite 4.0 Micro |
| LMArena Text: science | Not measured | 1055 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: medicine and healthcare | Not measured | 1009 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: legal and government | Not measured | 1071 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: business and finance | Not measured | 1000 | Only Llama 3.2 1B Instruct measured |
| Neither model has a result on the other 15 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| IFBench | 24.8% | 22.8% | Granite 4.0 Micro |
| Long context reasoning (AA-LCR) | 7.0% | 6.7% | Granite 4.0 Micro |
| LMArena Text | Not measured | 1055 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: instruction following | Not measured | 1031 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: creative writing | Not measured | 1033 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: English | Not measured | 1117 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: non-English | Not measured | 973 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: Chinese | Not measured | 959 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: German | Not measured | 1014 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: Russian | Not measured | 941 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: multi-turn | Not measured | 1030 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: longer queries | Not measured | 1050 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: writing and language | Not measured | 1030 | Only Llama 3.2 1B Instruct measured |
| LMArena Text: entertainment and media | Not measured | 1025 | Only Llama 3.2 1B Instruct measured |
| Neither model has a result on the other 34 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Granite 4.0 Micro | Llama 3.2 1B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Granite 4.0 Micro costs less
Repository review
50K input + 3K output tokens
Granite 4.0 Micro costs less
Agent loop
200K input + 10K output tokens
Granite 4.0 Micro costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | IBM Granite 4.0 Micro | Meta Llama 3.2 1B Instruct |
|---|---|---|
| Maker | IBM Granite | Meta Llama |
| Weights | Open weights | Open weights |
| Released | October 20, 2025 | September 25, 2024 |
| Context window | 131K | 60K |
| Input $/Mtok | $0.017 | $0.027 |
| Output $/Mtok | $0.112 | $0.201 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Granite 4.0 Micro better than Llama 3.2 1B Instruct for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Granite 4.0 Micro for output tokens 1.8x cheaper ($0.112 vs $0.201 per million) and a larger context window (131K against 60K tokens).
Which is cheaper, Granite 4.0 Micro or Llama 3.2 1B Instruct?
Granite 4.0 Micro costs $0.017 per million input tokens and $0.112 per million output tokens; Llama 3.2 1B Instruct costs $0.027 and $0.201. Cost per solved task is known for neither of them.
Which has the bigger context window, Granite 4.0 Micro or Llama 3.2 1B Instruct?
Granite 4.0 Micro, with 131K tokens against 60K.
How many benchmarks is the Granite 4.0 Micro vs Llama 3.2 1B Instruct comparison based on?
13 benchmarks have a result for both models as of October 22, 2026. Granite 4.0 Micro's coding index rests on 0 independent benchmarks and Llama 3.2 1B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.