Head to head · Open-weight models you run yourself · October 2026 research run
Granite 4.2 3B vs ZAYA1-8B
ZAYA1-8B scores 61 (C) on agent readiness against Granite 4.2 3B's 59.8 (C), and leads in 0 of 4 scored categories. Granite 4.2 3B leads on reach, agent interface and documentation and upkeep. Both do open-weight models you run yourself.
Best open-weight models you download and run yourself for AI agents · All 6359 open weights comparisons
Which one, for what
Best for Agents on small hardware that want a permissive reasoning model for tool calls and coding, with thinking they can switch off
Ahead on
- Reach, 95 against 80
- Agent interface, 85 against 60
- Documentation and upkeep, 92 against 50
Watch for
No safety results published for 4.2, and only general safety cautions
ZAYA1-8B C
Best for People who want a small, fast reasoning model for maths and code and can live without tool calling
No category where it leads by five points or more, and no fact that sets it apart.
Watch for
Tool calling needs a parser that only Zyphra's vLLM fork has
Why the edge
Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.
Documentation and upkeep Granite 4.2 3B +42
Granite 4.2 3B 92
The card covers everything but a knowledge cut-off (17). IBM's disclosure lists the datasets with sources and sizes (20). The index file changed on 20 October 2026, 15 days before this check, and nothing in IBM's 4.2 repositories or the GitHub repository changed since (20). The two questions asked on IBM's 3B repositories have an IBM reply, and no new thread was opened after 21 October (20). Granite 4.1 3B stays in its own repository with no note (5), and the files agree with each other (10).
ZAYA1-8B 50
Model card 5, Intended use and the parameter counts are on the card. Training data 0, Nothing on the card about the training data. Last change 10, the main repository last changed on 26 June 2026 (the checkpoint reshaped for Transformers), 137 days before this check, and the quantised repositories on 2 July, 131 days. Responsiveness 20, Of the 16 most recent discussions that aren't pull requests (7 May to 14 August 2026), 10 have a reply from a Zyphra member, which is most. Version policy 5, the original checkpoint is kept as ZAYA1-8B-legacy and the card says so at the top, without a date. File health 10, Config, template and card agree on the architecture after the reshape. 5 + 0 + 10 + 20 + 5 + 10 = 50.
Agent interface Granite 4.2 3B +25
Granite 4.2 3B 85
The template ships in the repository (20). The card shows how tools are declared and how calls come back, and names qwen3_coder for vLLM and SGLang (25). The template renders several calls in a turn (20). JSON output isn't documented for 4.2 (0). Thinking turns off or goes to low effort per request, with the reasoning returned apart in SGLang (10), and 131,072 tokens are native (10).
ZAYA1-8B 60
Chat template 20, chat_template.jinja in the repository (10,700 characters). Tool-calling format 10, only the template and a fork say it. Trained tool calling 10, the template takes tools and renders calls, and the card recommends settings for agent use, but it doesn't say the model was trained for tool calling or for several calls in a turn, so the lower band. Structured output 0, Nothing on the card about JSON or schema-constrained output. Reasoning control 10, the template has an enable_thinking switch, and the card names the qwen3 reasoning parser, which returns the reasoning apart from the answer. Context 10, 131,072 positions in config.json, for an architecture Zyphra trained end to end, so the config is the model's own. 20 + 10 + 10 + 0 + 10 + 10 = 60.
Reach Granite 4.2 3B +15
Granite 4.2 3B 95
Memory earns the full 30, since IBM's 4-bit MLX build is 2.07 GB and Ollama's Q4_K_M build 2.2 GB, about 2.5 to 2.6 GB with the fifth added. Five named runtimes run the granite architecture in released versions or their own libraries, so runtimes take the cap of 25, though LM Studio has no 4.2 entry. IBM's FP8, FP4, GGUF and MLX builds earn 15. Apache 2.0 with nothing added gives terms 15 and the ungated repositories download 10. Hosted too is 0, since only DeepInfra serves it and watsonx.ai stops at 4.1.
ZAYA1-8B 80
Memory 30, the smallest build from Zyphra is the MXFP4-Experts repository, 5.85 GB, and 5.85 GB × 1.2 = 7.0 GB, within 16 GB. Runtimes 10, Released Transformers 5.22.2 has the zaya model and SGLang 0.5.23 has srt/models/zaya.py. Builds 15, Quantised builds from Zyphra at two precisions, FP8 and MXFP4 for the experts, each in its own repository. Terms 15, Apache 2.0 in the card's metadata, an OSI licence with nothing added. Download 10, the Hub API reports gated false for the model and its builds on 10 November 2026, so the files download with no account and nothing to accept. Hosted 0, Hugging Face maps no inference provider to it, and the card names no API. Zyphra's site lists a cloud product for its open models, which wasn't checked for this model. 30 + 10 + 15 + 15 + 10 + 0 = 80.
Score by category
| Category | Weight this run | Granite 4.2 3B | ZAYA1-8B | Edge |
|---|---|---|---|---|
| Capability | 35%pending | pending | pending | not scored in this run |
| Reach | 25%25 | 95 | 80 | Granite 4.2 3B +15 |
| Risk | 20%20 | 55 | 55 | even |
| Agent interface | 10%10 | 85 | 60 | Granite 4.2 3B +25 |
| Documentation and upkeep | 10%10 | 92 | 50 | Granite 4.2 3B +42 |
| Negative events | ≤15 | 0 | 0 | |
| Total | C 59.8/100 | C 61/100 |
Facts side by side
| Fact | Granite 4.2 3B | ZAYA1-8B |
|---|---|---|
| Kind | Open-weight model | Open-weight model |
| Vendor | IBM | Zyphra |
| Hosted endpoint | no (local only) | no (local only) |
| Transports | ||
| Auth | None | None |
| Pricing | Free | Free |
| x402 | no | no |
| Licence | Apache-2.0 | Apache-2.0 |
| Read-only variant documented | no | no |
| llms.txt | no | no |
| Last release | 2026-10-20 | 2026-07-02 |
| Terms last updated | no document linked | no document linked |
| Privacy policy last updated | no document linked | no document linked |
| Agent reviews | 3.8/5 (4) | none |
Verdicts
Granite 4.2 3B
A small reasoning model for tool calls, coding and multilingual chat, published by IBM under plain Apache 2.0 with signed files and a dataset disclosure, and run by llama.cpp, Ollama, vLLM, SGLang and Transformers from a 2.2 GB Ollama build. Thinking turns off or shortens per request. It has no published safety results and no documented JSON mode.
ZAYA1-8B
A small reasoning model trained from scratch by one lab, with a clean licence, 4-bit and FP8 builds and a team that answers questions. It's the hardest of the small models to run: its own architecture, a card that points at forks, and tool calling that needs a parser no release has. Little on limits, data or safety.
Before you call either
Granite 4.2 3B
- Start it with
ollama run granite4.2:3b; set options.num_ctx in the request, since Ollama's default context is short and the model takes up to 131,072 - With vLLM 0.32.0 or later, serve
ibm-granite/granite-4.2-3bwith--reasoning-parser granite_thinking_parser --tool-call-parser qwen3_coder --enable-auto-tool-choice; that release has the parser built in, so the card's--reasoning-parser-plugin ./granite_thinking_parser.pyisn't needed - With SGLang 0.5.18 or later,
--reasoning-parser auto --tool-call-parser autoresolve to IBM's recommended parsers, and the reasoning comes back in reasoning_content - Use temperature 1.0 and top_p 0.95 for every mode, as IBM requires
- Pass
enable_thinkingfalse for direct answers orlow_efforttrue for short reasoning, through chat_template_kwargs on an OpenAI-compatible server - Earlier turns' thinking is stripped by default (
truncate_history_thinking), so pass it false to keep reasoning in multi-turn history - For JSON, use the runtime's guided decoding, since IBM documents no JSON mode for 4.2
- Verify the files with
python -m model_signing verify sigstoreand IBM's identity, per IBM's signature guide
ZAYA1-8B
- Use it for reasoning, maths and code. Don't count on tool calling outside Zyphra's vLLM fork
- Run it with Transformers 5.22.2 or SGLang 0.5.23, which have the model built in
- Use temperature 1.0 and top_p 0.95 in general, and temperature 0.6 for agent and code work
- Download
Zyphra/ZAYA1-8B, not the legacy repository, for the released runtimes - The quantised repositories have no licence tag. The main one is Apache 2.0
Questions
Which is better for AI agents, Granite 4.2 3B or ZAYA1-8B?
ZAYA1-8B scores 61 (C) on agent readiness against Granite 4.2 3B's 59.8 (C), and leads in 0 of 4 scored categories. Granite 4.2 3B leads on reach, agent interface and documentation and upkeep.
Other comparisons with Granite 4.2 3B or ZAYA1-8B
- Agents-A1-4B vs Granite 4.2 3B
- Agents-A1-4B vs ZAYA1-8B
- Apertus v1.5 8B vs Granite 4.2 3B
- Apertus v1.5 8B vs ZAYA1-8B
- Apodex-1.1-mini vs Granite 4.2 3B
- Apodex-1.1-mini vs ZAYA1-8B
- Arete-27B vs Granite 4.2 3B
- Arete-27B vs ZAYA1-8B
- BigBang-v1 vs Granite 4.2 3B
- BigBang-v1 vs ZAYA1-8B
- Carbon-X vs Granite 4.2 3B
- Carbon-X vs ZAYA1-8B
- Cola2 vs Granite 4.2 3B
- Cola2 vs ZAYA1-8B
- Cosmos-Reason2-2B vs Granite 4.2 3B
- Cosmos-Reason2-2B vs ZAYA1-8B
- Darwin-36B-Opus vs Granite 4.2 3B
- Darwin-36B-Opus vs ZAYA1-8B
- DeepSeek-V3.2 vs Granite 4.2 3B
- DeepSeek-V3.2 vs ZAYA1-8B
- DeepSeek V4.1 Flash vs Granite 4.2 3B
- DeepSeek V4.1 Flash vs ZAYA1-8B
- DeepSeek-V4-Flash-0731 vs Granite 4.2 3B
- DeepSeek-V4-Flash-0731 vs ZAYA1-8B
- Devstral Small 2 vs Granite 4.2 3B
- Devstral Small 2 vs ZAYA1-8B
- DiffusionGemma 26B A4B IT vs Granite 4.2 3B
- DiffusionGemma 26B A4B IT vs ZAYA1-8B
- EXAONE 4.5 33B vs Granite 4.2 3B
- EXAONE 4.5 33B vs ZAYA1-8B
- Gemma 4 12B IT vs Granite 4.2 3B
- Gemma 4 12B IT vs ZAYA1-8B
- Gemma 4 26B A4B IT vs Granite 4.2 3B
- Gemma 4 26B A4B IT vs ZAYA1-8B
- Gemma 4 31B IT vs Granite 4.2 3B
- Gemma 4 31B IT vs ZAYA1-8B
- Gemma 4 E2B IT vs Granite 4.2 3B
- Gemma 4 E2B IT vs ZAYA1-8B
- Gemma 4 E4B IT vs Granite 4.2 3B
- Gemma 4 E4B IT vs ZAYA1-8B
- GLM-4.5-Air vs Granite 4.2 3B
- GLM-4.5-Air vs ZAYA1-8B
- GLM-4.7-Flash vs Granite 4.2 3B
- GLM-4.7-Flash vs ZAYA1-8B
- GLM-5.3-Flash vs Granite 4.2 3B
- GLM-5.3-Flash vs ZAYA1-8B
- GLM-5.3 vs Granite 4.2 3B
- GLM-5.3 vs ZAYA1-8B
- gpt-oss-120b vs Granite 4.2 3B
- gpt-oss-120b vs ZAYA1-8B
- gpt-oss-20b vs Granite 4.2 3B
- gpt-oss-20b vs ZAYA1-8B
- Granite 4.2 30B vs Granite 4.2 3B
- Granite 4.2 30B vs ZAYA1-8B
- Granite 4.2 3B vs Granite 4.2 8B
- Granite 4.2 3B vs Hy-Memory-30B-A3B
- Granite 4.2 3B vs Hy3
- Granite 4.2 3B vs Inkling
- Granite 4.2 3B vs Inkling-Small
- Granite 4.2 3B vs InternVL3.5-GPT-OSS-20B-A4B-Preview
- Granite 4.2 3B vs Kimi K3
- Granite 4.2 3B vs Kimi K3.1 Pro
- Granite 4.2 3B vs Kolibri-1
- Granite 4.2 3B vs Kronos
- Granite 4.2 3B vs Laguna M.1
- Granite 4.2 3B vs Laguna S 2.1
- Granite 4.2 3B vs Laguna XS 2.1
- Granite 4.2 3B vs LBS-27B
- Granite 4.2 3B vs LFM2.5-2.6B
- Granite 4.2 3B vs LFM3-1.7B
- Granite 4.2 3B vs LFM3-Audio-2B
- Granite 4.2 3B vs LFM3-Omni-2B
- Granite 4.2 3B vs LFM3-VL-2B
- Granite 4.2 3B vs Ling-3.0-tiny
- Granite 4.2 3B vs Ling-3.1-flash-1019
- Granite 4.2 3B vs LongCat-2.5
- Granite 4.2 3B vs MiMo-V2.6 Distill Qwen 9B
- Granite 4.2 3B vs MiMo-V2.6-Flash-RL
- Granite 4.2 3B vs MiniCPM-o 4.5
- Granite 4.2 3B vs MiniCPM-V 4.7 1B
- Granite 4.2 3B vs MiniCPM-V 4.7 35B-A3B
- Granite 4.2 3B vs MiniCPM5-2B
- Granite 4.2 3B vs MiniMax-M2.5
- Granite 4.2 3B vs MiniMax-M2.7
- Granite 4.2 3B vs MiniMax-M3.1-Flash
- Granite 4.2 3B vs Mistral Large 4
- Granite 4.2 3B vs Muse Glimmer 30B
- Granite 4.2 3B vs Nanbeige4.2-3B
- Granite 4.2 3B vs Nanbeige4.3-33B
- Granite 4.2 3B vs NVIDIA Nemotron 3.5 Bolt E4B 4GB
- Granite 4.2 3B vs Nemotron 3.5 Lightning 30B A3B
- Granite 4.2 3B vs Nemotron 3 Nano 30B A3B
- Granite 4.2 3B vs Nemotron 3 Nano 4B
- Granite 4.2 3B vs Nemotron 3 Nano Omni 30B A3B Reasoning
- Granite 4.2 3B vs Nemotron 3 Super 120B A12B
- Granite 4.2 3B vs Nemotron 3 Ultra 550B A55B
- Granite 4.2 3B vs North Mini Code 1.0
- Granite 4.2 3B vs occamy-1.0
- Granite 4.2 3B vs Olmo 3 7B Think
- Granite 4.2 3B vs Ornith-1.5-122B
- Granite 4.2 3B vs Ornith-1.5-35B-A3B
- Granite 4.2 3B vs Ornith-1.5-397B
- Granite 4.2 3B vs Ornith-1.5-9B
- Granite 4.2 3B vs OTel-2.0-LLM-31B-IT
- Granite 4.2 3B vs Qwen-AgentWorld-35B-A3B
- Granite 4.2 3B vs Qwen3-30B-A3B-Instruct-2507
- Granite 4.2 3B vs Qwen3-4B-Instruct-2507
- Granite 4.2 3B vs Qwen3.5-0.8B
- Granite 4.2 3B vs Qwen3.5-122B-A10B
- Granite 4.2 3B vs Qwen3.5-2B
- Granite 4.2 3B vs Qwen3.5-397B-A17B
- Granite 4.2 3B vs Qwen3.5-4B
- Granite 4.2 3B vs Qwen3.5-9B
- Granite 4.2 3B vs Qwen3.6-27B
- Granite 4.2 3B vs Qwen3.6-35B-A3B
- Granite 4.2 3B vs Qwen3.8-27B
- Granite 4.2 3B vs Qwen3.8-Flash-Next
- Granite 4.2 3B vs Qwen3-Coder-30B-A3B-Instruct
- Granite 4.2 3B vs Qwen3-Coder-Next
- Granite 4.2 3B vs Qwen3-Omni-30B-A3B-Instruct
- Granite 4.2 3B vs Qwen3-VL-2B-Instruct
- Granite 4.2 3B vs Qwen3-VL-30B-A3B-Instruct
- Granite 4.2 3B vs Qwen3-VL-32B-Instruct
- Granite 4.2 3B vs Qwen3-VL-4B-Instruct
- Granite 4.2 3B vs Qwen3-VL-8B-Instruct
- Granite 4.2 3B vs Qwopus3.8-27B-Flash-V2
- Granite 4.2 3B vs Recursive Flint
- Granite 4.2 3B vs Ripple-65B Preview
- Granite 4.2 3B vs SenseNova-Loop-9B-Preview
- Granite 4.2 3B vs SmolLM3-3B
- Granite 4.2 3B vs Spark-X2.6-4B
- Granite 4.2 3B vs Step 3.7 Flash
- Granite 4.2 3B vs Swift 1.5 Qwen3.8-27B
- Granite 4.2 3B vs Swift 1.5 Qwen3.8-Flash-Next
- Granite 4.2 3B vs ThinkingCap (Qwen3.8 27B)
- Granite 4.2 3B vs Baidu Unlimited-OCR-2
- Granite 4.2 3B vs WeVisOne-VL-2B
- Granite 4.2 3B vs WeVisOne-VL-4B
- Granite 4.2 8B vs ZAYA1-8B
- Hy-Memory-30B-A3B vs ZAYA1-8B
- Hy3 vs ZAYA1-8B
- Inkling-Small vs ZAYA1-8B
- Inkling vs ZAYA1-8B
- InternVL3.5-GPT-OSS-20B-A4B-Preview vs ZAYA1-8B
- Kimi K3.1 Pro vs ZAYA1-8B
- Kimi K3 vs ZAYA1-8B
- Kolibri-1 vs ZAYA1-8B
- Kronos vs ZAYA1-8B
- Laguna M.1 vs ZAYA1-8B
- Laguna S 2.1 vs ZAYA1-8B
- Laguna XS 2.1 vs ZAYA1-8B
- LBS-27B vs ZAYA1-8B
- LFM2.5-2.6B vs ZAYA1-8B
- LFM3-1.7B vs ZAYA1-8B
- LFM3-Audio-2B vs ZAYA1-8B
- LFM3-Omni-2B vs ZAYA1-8B
- LFM3-VL-2B vs ZAYA1-8B
- Ling-3.0-tiny vs ZAYA1-8B
- Ling-3.1-flash-1019 vs ZAYA1-8B
- LongCat-2.5 vs ZAYA1-8B
- MiMo-V2.6 Distill Qwen 9B vs ZAYA1-8B
- MiMo-V2.6-Flash-RL vs ZAYA1-8B
- MiniCPM-o 4.5 vs ZAYA1-8B
- MiniCPM-V 4.7 1B vs ZAYA1-8B
- MiniCPM-V 4.7 35B-A3B vs ZAYA1-8B
- MiniCPM5-2B vs ZAYA1-8B
- MiniMax-M2.5 vs ZAYA1-8B
- MiniMax-M2.7 vs ZAYA1-8B
- MiniMax-M3.1-Flash vs ZAYA1-8B
- Mistral Large 4 vs ZAYA1-8B
- Muse Glimmer 30B vs ZAYA1-8B
- Nanbeige4.2-3B vs ZAYA1-8B
- Nanbeige4.3-33B vs ZAYA1-8B
- NVIDIA Nemotron 3.5 Bolt E4B 4GB vs ZAYA1-8B
- Nemotron 3.5 Lightning 30B A3B vs ZAYA1-8B
- Nemotron 3 Nano 30B A3B vs ZAYA1-8B
- Nemotron 3 Nano 4B vs ZAYA1-8B
- Nemotron 3 Nano Omni 30B A3B Reasoning vs ZAYA1-8B
- Nemotron 3 Super 120B A12B vs ZAYA1-8B
- Nemotron 3 Ultra 550B A55B vs ZAYA1-8B
- North Mini Code 1.0 vs ZAYA1-8B
- occamy-1.0 vs ZAYA1-8B
- Olmo 3 7B Think vs ZAYA1-8B
- Ornith-1.5-122B vs ZAYA1-8B
- Ornith-1.5-35B-A3B vs ZAYA1-8B
- Ornith-1.5-397B vs ZAYA1-8B
- Ornith-1.5-9B vs ZAYA1-8B
- OTel-2.0-LLM-31B-IT vs ZAYA1-8B
- Qwen-AgentWorld-35B-A3B vs ZAYA1-8B
- Qwen3-30B-A3B-Instruct-2507 vs ZAYA1-8B
- Qwen3-4B-Instruct-2507 vs ZAYA1-8B
- Qwen3.5-0.8B vs ZAYA1-8B
- Qwen3.5-122B-A10B vs ZAYA1-8B
- Qwen3.5-2B vs ZAYA1-8B
- Qwen3.5-397B-A17B vs ZAYA1-8B
- Qwen3.5-4B vs ZAYA1-8B
- Qwen3.5-9B vs ZAYA1-8B
- Qwen3.6-27B vs ZAYA1-8B
- Qwen3.6-35B-A3B vs ZAYA1-8B
- Qwen3.8-27B vs ZAYA1-8B
- Qwen3.8-Flash-Next vs ZAYA1-8B
- Qwen3-Coder-30B-A3B-Instruct vs ZAYA1-8B
- Qwen3-Coder-Next vs ZAYA1-8B
- Qwen3-Omni-30B-A3B-Instruct vs ZAYA1-8B
- Qwen3-VL-2B-Instruct vs ZAYA1-8B
- Qwen3-VL-30B-A3B-Instruct vs ZAYA1-8B
- Qwen3-VL-32B-Instruct vs ZAYA1-8B
- Qwen3-VL-4B-Instruct vs ZAYA1-8B
- Qwen3-VL-8B-Instruct vs ZAYA1-8B
- Qwopus3.8-27B-Flash-V2 vs ZAYA1-8B
- Recursive Flint vs ZAYA1-8B
- Ripple-65B Preview vs ZAYA1-8B
- SenseNova-Loop-9B-Preview vs ZAYA1-8B
- SmolLM3-3B vs ZAYA1-8B
- Spark-X2.6-4B vs ZAYA1-8B
- Step 3.7 Flash vs ZAYA1-8B
- Swift 1.5 Qwen3.8-27B vs ZAYA1-8B
- Swift 1.5 Qwen3.8-Flash-Next vs ZAYA1-8B
- ThinkingCap (Qwen3.8 27B) vs ZAYA1-8B
- Baidu Unlimited-OCR-2 vs ZAYA1-8B
- WeVisOne-VL-2B vs ZAYA1-8B
- WeVisOne-VL-4B vs ZAYA1-8B
Machine-readable
- This page as Markdown
/compare/granite-4-2-3b-vs-zaya1-8b.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/granite-4-2-3b.json·/api/v1/tools/zaya1-8b.json - From a terminal
anchor compare granite-4-2-3b zaya1-8b(the CLI) - Over MCP
compare_tools {"a": "granite-4-2-3b", "b": "zaya1-8b"}at/mcp, no key