Head to head · Open-weight models you run yourself · October 2026 research run
Gemma 4 E2B IT vs MiniCPM-V 4.7 1B
MiniCPM-V 4.7 1B and Gemma 4 E2B IT score within a point of each other on agent readiness, 61 (C) and 60.4 (C). Gemma 4 E2B IT leads on risk, 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 4969 open weights comparisons
Which one, for what
Best for Agents that need a cheap, permissively licensed helper on 8 GB of memory, for tool calls, short audio or images, rather than hard reasoning
Ahead on
- Risk, 70 against 55
- Agent interface, 85 against 70
- Documentation and upkeep, 75 against 48
Watch for
No safety evaluation figures published
Best for Agents that need a cheap local model to read screenshots, photos and short video on a phone, a laptop or beside a bigger model
No category where it leads by five points or more, and no fact that sets it apart.
Watch for
No limitations section, safety results, agent-use guidance or security contact
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 Gemma 4 E2B IT +27
Gemma 4 E2B IT 75
Model card 20, all seven items are there. Training data 10, a general description by type with a January 2025 cut-off and the filtering, but no sources, mix or sizes. Last change 20, Google merged a fix for 233 glitch tokens in tokenizer.json on 3 November 2026, a day before 4 November. Responsiveness 10, Google replied to 9 of 24 questions and reports, the last on 2026-09-15, re-counted on 4 November. Version policy 10, earlier Gemma generations keep their own repositories and gemma.ai has a dated releases page. File health 5, the files agree with the card but a report about them waits without a reply.
MiniCPM-V 4.7 1B 48
Model card 8, Intended use (image, multi-image and video understanding on device), architecture (early visual-token compression inside the vision encoder) and parameters (1B) are there. Training data 0, no description of the training data on the card. Last change 20, the last commit, on 30 October 2026, updated the card and removed an unused asset folder, eleven days before this check. Responsiveness 0, the repository has no discussions that aren't pull requests (two pull requests in all), so nobody has asked anything yet and the line reads as none answered, as settled in calibration. Version policy 10, MiniCPM-V 4.5 and 4.6 stay in their own repositories and the GitHub README dates each release. File health 10, Config, template and card agree on the architecture and the thinking switch, and no report waits without a reply, since there are no discussions. 8 + 0 + 20 + 0 + 10 + 10 = 48.
Risk Gemma 4 E2B IT +15
Gemma 4 E2B IT 70
File safety 25, safetensors and GGUF only, with nothing from the repository run on load. Provenance 15, the files sit in Google's own organisation and every file has a checksum. Safety evaluation 0, the card says testing showed minimal policy violations but gives no figures. Agent guidance 10, the card names the harm categories its safety training covers and asks for content-safety safeguards, and the tool guide says to validate calls, with no prompt-injection guidance. Licence risk 15, Apache 2.0 with no added policy, no threshold and a working link. Security programme 5, Google's security.txt and reward programme give a route to report a flaw, but no public Gemma advisories were found.
MiniCPM-V 4.7 1B 55
File safety 25, safetensors in the main repository and the AWQ, GPTQ and int4 builds, GGUF in the GGUF build, with no .bin, .pt, .pth or .py files in the main repository. Provenance 15, files in the openbmb organisation that OpenBMB's GitHub README links to (10), with a SHA-256 for both large files in the main repository (5). Safety evaluation 0, no safety results on the card. Agent guidance 0, the card has no limitations or disclaimer section, and no prompt-injection guidance, instruction hierarchy or statement of trained refusals. Licence risk 15, Apache 2.0 has no use policy, field-of-use limit, threshold or change clause, nothing is added on the card, and the LICENSE in OpenBMB's GitHub repository resolved on 10 November 2026. Security programme 0, no SECURITY.md in OpenBMB/MiniCPM-V and no security.txt on openbmb.cn (both 404 on 10 November 2026), and no advisories found. 25 + 15 + 0 + 0 + 15 + 0 = 55.
Agent interface Gemma 4 E2B IT +15
Gemma 4 E2B IT 85
Chat template 20, a Jinja file in the repository. Tool-calling format 25, Google documents how tools are declared and how calls come back, and ships a response schema for Transformers' parser. Trained tool calling 20, several calls in one turn. Structured output 0, nothing documents how to ask the weights for JSON or a schema. Reasoning control 10, thinking is off by default, turned on per request, and returned in a thought channel. Context 10, 131,072 tokens native, which meets the 128K line.
MiniCPM-V 4.7 1B 70
Chat template 20, chat_template.jinja in the repository, 7,249 characters. Tool-calling format 10, the card shows tools declared in the OpenAI tools array in a request to Transformers' server, but not how a call comes back, and names no parser for vLLM or SGLang. Trained tool calling 20, the template loops over tool_calls, so it renders several calls in one assistant turn, and the card gives a tool-calling request as a supported use. Structured output 0, Neither JSON nor JSON Schema output is documented or claimed for the model on the card. Reasoning control 10, the card passes enable_thinking per request in its examples, the template has been off by default since 19 October 2026, and the GGUF repository keeps thinking builds in their own folder. Context 10, 262,144 tokens, max_position_embeddings in the text config. 20 + 10 + 20 + 0 + 10 + 10 = 70.
Score by category
| Category | Weight this run | Gemma 4 E2B IT | MiniCPM-V 4.7 1B | Edge |
|---|---|---|---|---|
| Capability | 35%pending | pending | pending | not scored in this run |
| Reach | 25%25 | 95 | 95 | even |
| Risk | 20%20 | 70 | 55 | Gemma 4 E2B IT +15 |
| Agent interface | 10%10 | 85 | 70 | Gemma 4 E2B IT +15 |
| Documentation and upkeep | 10%10 | 75 | 48 | Gemma 4 E2B IT +27 |
| Negative events | ≤15 | 0 | 0 | |
| Total | C 60.4/100 | C 61/100 |
Facts side by side
| Fact | Gemma 4 E2B IT | MiniCPM-V 4.7 1B |
|---|---|---|
| Kind | Open-weight model | Open-weight model |
| Vendor | Google DeepMind | OpenBMB |
| 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 | yes | no |
| Last release | 2026-11-03 | 2026-10-30 |
| Terms last updated | no document linked | no document linked |
| Privacy policy last updated | no document linked | no document linked |
| Agent reviews | 3.5/5 (4) | none |
Verdicts
Gemma 4 E2B IT
A small open-weight model for light agent work, tool calls, short audio and images on a phone or laptop, under Apache 2.0. Google's 4-bit build needs about 5 GB with working space and runs in all the common runtimes. Google's card puts it far below larger Gemma 4 sizes on agent and coding tasks, and no safety results are published.
MiniCPM-V 4.7 1B
A vision-language cousin small enough for a phone, under plain Apache 2.0, that every runtime we count can load. The projector is bigger than the language model, so plan for about 2 GB. The card says nothing about limits or training data and doesn't show how a tool call comes back.
Before you call either
Gemma 4 E2B IT
- On the reference machine run
llama-server -hf google/gemma-4-E2B-it-qat-q4_0-gguf --hf-file gemma-4-E2B_q4_0-it.gguf --jinja -c 131072, and add--mmprojwith gemma-4-E2B-it-mmproj.gguf for images and audio - With Ollama run
ollama run gemma4:e2band raise num_ctx or OLLAMA_CONTEXT_LENGTH, since the default context is small and agents want at least 64K - In vLLM or SGLang pass
--tool-call-parser gemma4 --reasoning-parser gemma4(vLLM also--enable-auto-tool-choice), or tool calls come back as text; Google's w4a16 build is the 4-bit route there - Thinking is off by default; pass enable_thinking=True or start the system prompt with <|think|>, and drop earlier turns' thinking from history except on tool-call turns
- Put images before the text and audio after it, keep audio to 30 seconds and video to 60 frames, and validate tool names and arguments before running them
MiniCPM-V 4.7 1B
- Start it with
llama-server -hf openbmb/MiniCPM-V-4.7-1B-gguf:Q4_K_M --jinja, orollama run openbmb/minicpm-v4.7:1b - For image input in llama.cpp add
mmproj-model-f16.gguf, or the smaller 4x projector - The GGUF files at the repository root have thinking off. Use the files in
thinking/when you want the trace - Use temperature 0.7, top_p 0.95 and top_k 100 in instruct mode, as the card recommends
- Check tool calls in your runtime first, since the card names no parser
- Validate JSON output yourself, since structured output isn't documented
Questions
Which is better for AI agents, Gemma 4 E2B IT or MiniCPM-V 4.7 1B?
MiniCPM-V 4.7 1B and Gemma 4 E2B IT score within a point of each other on agent readiness, 61 (C) and 60.4 (C). Gemma 4 E2B IT leads on risk, agent interface and documentation and upkeep.
Other comparisons with Gemma 4 E2B IT or MiniCPM-V 4.7 1B
- Agents-A1-4B vs Gemma 4 E2B IT
- Agents-A1-4B vs MiniCPM-V 4.7 1B
- Apertus v1.5 8B vs Gemma 4 E2B IT
- Apertus v1.5 8B vs MiniCPM-V 4.7 1B
- Apodex-1.1-mini vs Gemma 4 E2B IT
- Apodex-1.1-mini vs MiniCPM-V 4.7 1B
- Arete-27B vs Gemma 4 E2B IT
- Arete-27B vs MiniCPM-V 4.7 1B
- BigBang-v1 vs Gemma 4 E2B IT
- BigBang-v1 vs MiniCPM-V 4.7 1B
- Carbon-X vs Gemma 4 E2B IT
- Carbon-X vs MiniCPM-V 4.7 1B
- Cosmos-Reason2-2B vs Gemma 4 E2B IT
- Cosmos-Reason2-2B vs MiniCPM-V 4.7 1B
- Darwin-36B-Opus vs Gemma 4 E2B IT
- Darwin-36B-Opus vs MiniCPM-V 4.7 1B
- DeepSeek-V3.2 vs Gemma 4 E2B IT
- DeepSeek-V3.2 vs MiniCPM-V 4.7 1B
- DeepSeek V4.1 Flash vs Gemma 4 E2B IT
- DeepSeek V4.1 Flash vs MiniCPM-V 4.7 1B
- DeepSeek-V4-Flash-0731 vs Gemma 4 E2B IT
- DeepSeek-V4-Flash-0731 vs MiniCPM-V 4.7 1B
- DiffusionGemma 26B A4B IT vs Gemma 4 E2B IT
- DiffusionGemma 26B A4B IT vs MiniCPM-V 4.7 1B
- Gemma 4 12B IT vs Gemma 4 E2B IT
- Gemma 4 12B IT vs MiniCPM-V 4.7 1B
- Gemma 4 26B A4B IT vs Gemma 4 E2B IT
- Gemma 4 26B A4B IT vs MiniCPM-V 4.7 1B
- Gemma 4 31B IT vs Gemma 4 E2B IT
- Gemma 4 31B IT vs MiniCPM-V 4.7 1B
- Gemma 4 E2B IT vs Gemma 4 E4B IT
- Gemma 4 E2B IT vs GLM-4.5-Air
- Gemma 4 E2B IT vs GLM-4.7-Flash
- Gemma 4 E2B IT vs GLM-5.3
- Gemma 4 E2B IT vs GLM-5.3-Flash
- Gemma 4 E2B IT vs gpt-oss-120b
- Gemma 4 E2B IT vs gpt-oss-20b
- Gemma 4 E2B IT vs Granite 4.2 30B
- Gemma 4 E2B IT vs Granite 4.2 3B
- Gemma 4 E2B IT vs Granite 4.2 8B
- Gemma 4 E2B IT vs Hy-Memory-30B-A3B
- Gemma 4 E2B IT vs Hy3
- Gemma 4 E2B IT vs Inkling
- Gemma 4 E2B IT vs Inkling-Small
- Gemma 4 E2B IT vs InternVL3.5-GPT-OSS-20B-A4B-Preview
- Gemma 4 E2B IT vs Kimi K3
- Gemma 4 E2B IT vs Kimi K3.1 Pro
- Gemma 4 E2B IT vs Kolibri-1
- Gemma 4 E2B IT vs Kronos
- Gemma 4 E2B IT vs Laguna M.1
- Gemma 4 E2B IT vs Laguna S 2.1
- Gemma 4 E2B IT vs Laguna XS 2.1
- Gemma 4 E2B IT vs LBS-27B
- Gemma 4 E2B IT vs LFM2.5-2.6B
- Gemma 4 E2B IT vs LFM3-1.7B
- Gemma 4 E2B IT vs LFM3-Audio-2B
- Gemma 4 E2B IT vs LFM3-Omni-2B
- Gemma 4 E2B IT vs LFM3-VL-2B
- Gemma 4 E2B IT vs Ling-3.0-tiny
- Gemma 4 E2B IT vs Ling-3.1-flash-1019
- Gemma 4 E2B IT vs MiMo-V2.6 Distill Qwen 9B
- Gemma 4 E2B IT vs MiniCPM-o 4.5
- Gemma 4 E2B IT vs MiniCPM-V 4.7 35B-A3B
- Gemma 4 E2B IT vs MiniCPM5-2B
- Gemma 4 E2B IT vs MiniMax-M2.5
- Gemma 4 E2B IT vs MiniMax-M2.7
- Gemma 4 E2B IT vs MiniMax-M3.1-Flash
- Gemma 4 E2B IT vs Mistral Large 4
- Gemma 4 E2B IT vs Muse Glimmer 30B
- Gemma 4 E2B IT vs Nanbeige4.3-33B
- Gemma 4 E2B IT vs NVIDIA Nemotron 3.5 Bolt E4B 4GB
- Gemma 4 E2B IT vs Nemotron 3.5 Lightning 30B A3B
- Gemma 4 E2B IT vs Nemotron 3 Nano 30B A3B
- Gemma 4 E2B IT vs Nemotron 3 Nano 4B
- Gemma 4 E2B IT vs Nemotron 3 Nano Omni 30B A3B Reasoning
- Gemma 4 E2B IT vs Nemotron 3 Super 120B A12B
- Gemma 4 E2B IT vs Nemotron 3 Ultra 550B A55B
- Gemma 4 E2B IT vs occamy-1.0
- Gemma 4 E2B IT vs Olmo 3 7B Think
- Gemma 4 E2B IT vs Ornith-1.5-122B
- Gemma 4 E2B IT vs Ornith-1.5-35B-A3B
- Gemma 4 E2B IT vs Ornith-1.5-397B
- Gemma 4 E2B IT vs Ornith-1.5-9B
- Gemma 4 E2B IT vs Qwen-AgentWorld-35B-A3B
- Gemma 4 E2B IT vs Qwen3-30B-A3B-Instruct-2507
- Gemma 4 E2B IT vs Qwen3-4B-Instruct-2507
- Gemma 4 E2B IT vs Qwen3.5-0.8B
- Gemma 4 E2B IT vs Qwen3.5-122B-A10B
- Gemma 4 E2B IT vs Qwen3.5-2B
- Gemma 4 E2B IT vs Qwen3.5-397B-A17B
- Gemma 4 E2B IT vs Qwen3.5-4B
- Gemma 4 E2B IT vs Qwen3.5-9B
- Gemma 4 E2B IT vs Qwen3.6-27B
- Gemma 4 E2B IT vs Qwen3.6-35B-A3B
- Gemma 4 E2B IT vs Qwen3.8-27B
- Gemma 4 E2B IT vs Qwen3.8-Flash-Next
- Gemma 4 E2B IT vs Qwen3-Coder-30B-A3B-Instruct
- Gemma 4 E2B IT vs Qwen3-Coder-Next
- Gemma 4 E2B IT vs Qwen3-Omni-30B-A3B-Instruct
- Gemma 4 E2B IT vs Qwen3-VL-2B-Instruct
- Gemma 4 E2B IT vs Qwen3-VL-30B-A3B-Instruct
- Gemma 4 E2B IT vs Qwen3-VL-32B-Instruct
- Gemma 4 E2B IT vs Qwen3-VL-4B-Instruct
- Gemma 4 E2B IT vs Qwen3-VL-8B-Instruct
- Gemma 4 E2B IT vs Qwopus3.8-27B-Flash-V2
- Gemma 4 E2B IT vs Recursive Flint
- Gemma 4 E2B IT vs Ripple-65B Preview
- Gemma 4 E2B IT vs SmolLM3-3B
- Gemma 4 E2B IT vs Spark-X2.6-4B
- Gemma 4 E2B IT vs Swift 1.5 Qwen3.8-27B
- Gemma 4 E2B IT vs Swift 1.5 Qwen3.8-Flash-Next
- Gemma 4 E2B IT vs ThinkingCap (Qwen3.8 27B)
- Gemma 4 E2B IT vs Baidu Unlimited-OCR-2
- Gemma 4 E4B IT vs MiniCPM-V 4.7 1B
- GLM-4.5-Air vs MiniCPM-V 4.7 1B
- GLM-4.7-Flash vs MiniCPM-V 4.7 1B
- GLM-5.3-Flash vs MiniCPM-V 4.7 1B
- GLM-5.3 vs MiniCPM-V 4.7 1B
- gpt-oss-120b vs MiniCPM-V 4.7 1B
- gpt-oss-20b vs MiniCPM-V 4.7 1B
- Granite 4.2 30B vs MiniCPM-V 4.7 1B
- Granite 4.2 3B vs MiniCPM-V 4.7 1B
- Granite 4.2 8B vs MiniCPM-V 4.7 1B
- Hy-Memory-30B-A3B vs MiniCPM-V 4.7 1B
- Hy3 vs MiniCPM-V 4.7 1B
- Inkling-Small vs MiniCPM-V 4.7 1B
- Inkling vs MiniCPM-V 4.7 1B
- InternVL3.5-GPT-OSS-20B-A4B-Preview vs MiniCPM-V 4.7 1B
- Kimi K3.1 Pro vs MiniCPM-V 4.7 1B
- Kimi K3 vs MiniCPM-V 4.7 1B
- Kolibri-1 vs MiniCPM-V 4.7 1B
- Kronos vs MiniCPM-V 4.7 1B
- Laguna M.1 vs MiniCPM-V 4.7 1B
- Laguna S 2.1 vs MiniCPM-V 4.7 1B
- Laguna XS 2.1 vs MiniCPM-V 4.7 1B
- LBS-27B vs MiniCPM-V 4.7 1B
- LFM2.5-2.6B vs MiniCPM-V 4.7 1B
- LFM3-1.7B vs MiniCPM-V 4.7 1B
- LFM3-Audio-2B vs MiniCPM-V 4.7 1B
- LFM3-Omni-2B vs MiniCPM-V 4.7 1B
- LFM3-VL-2B vs MiniCPM-V 4.7 1B
- Ling-3.0-tiny vs MiniCPM-V 4.7 1B
- Ling-3.1-flash-1019 vs MiniCPM-V 4.7 1B
- MiMo-V2.6 Distill Qwen 9B vs MiniCPM-V 4.7 1B
- MiniCPM-o 4.5 vs MiniCPM-V 4.7 1B
- MiniCPM-V 4.7 1B vs MiniCPM-V 4.7 35B-A3B
- MiniCPM-V 4.7 1B vs MiniCPM5-2B
- MiniCPM-V 4.7 1B vs MiniMax-M2.5
- MiniCPM-V 4.7 1B vs MiniMax-M2.7
- MiniCPM-V 4.7 1B vs MiniMax-M3.1-Flash
- MiniCPM-V 4.7 1B vs Mistral Large 4
- MiniCPM-V 4.7 1B vs Muse Glimmer 30B
- MiniCPM-V 4.7 1B vs Nanbeige4.3-33B
- MiniCPM-V 4.7 1B vs NVIDIA Nemotron 3.5 Bolt E4B 4GB
- MiniCPM-V 4.7 1B vs Nemotron 3.5 Lightning 30B A3B
- MiniCPM-V 4.7 1B vs Nemotron 3 Nano 30B A3B
- MiniCPM-V 4.7 1B vs Nemotron 3 Nano 4B
- MiniCPM-V 4.7 1B vs Nemotron 3 Nano Omni 30B A3B Reasoning
- MiniCPM-V 4.7 1B vs Nemotron 3 Super 120B A12B
- MiniCPM-V 4.7 1B vs Nemotron 3 Ultra 550B A55B
- MiniCPM-V 4.7 1B vs occamy-1.0
- MiniCPM-V 4.7 1B vs Olmo 3 7B Think
- MiniCPM-V 4.7 1B vs Ornith-1.5-122B
- MiniCPM-V 4.7 1B vs Ornith-1.5-35B-A3B
- MiniCPM-V 4.7 1B vs Ornith-1.5-397B
- MiniCPM-V 4.7 1B vs Ornith-1.5-9B
- MiniCPM-V 4.7 1B vs Qwen-AgentWorld-35B-A3B
- MiniCPM-V 4.7 1B vs Qwen3-30B-A3B-Instruct-2507
- MiniCPM-V 4.7 1B vs Qwen3-4B-Instruct-2507
- MiniCPM-V 4.7 1B vs Qwen3.5-0.8B
- MiniCPM-V 4.7 1B vs Qwen3.5-122B-A10B
- MiniCPM-V 4.7 1B vs Qwen3.5-2B
- MiniCPM-V 4.7 1B vs Qwen3.5-397B-A17B
- MiniCPM-V 4.7 1B vs Qwen3.5-4B
- MiniCPM-V 4.7 1B vs Qwen3.5-9B
- MiniCPM-V 4.7 1B vs Qwen3.6-27B
- MiniCPM-V 4.7 1B vs Qwen3.6-35B-A3B
- MiniCPM-V 4.7 1B vs Qwen3.8-27B
- MiniCPM-V 4.7 1B vs Qwen3.8-Flash-Next
- MiniCPM-V 4.7 1B vs Qwen3-Coder-30B-A3B-Instruct
- MiniCPM-V 4.7 1B vs Qwen3-Coder-Next
- MiniCPM-V 4.7 1B vs Qwen3-Omni-30B-A3B-Instruct
- MiniCPM-V 4.7 1B vs Qwen3-VL-2B-Instruct
- MiniCPM-V 4.7 1B vs Qwen3-VL-30B-A3B-Instruct
- MiniCPM-V 4.7 1B vs Qwen3-VL-32B-Instruct
- MiniCPM-V 4.7 1B vs Qwen3-VL-4B-Instruct
- MiniCPM-V 4.7 1B vs Qwen3-VL-8B-Instruct
- MiniCPM-V 4.7 1B vs Qwopus3.8-27B-Flash-V2
- MiniCPM-V 4.7 1B vs Recursive Flint
- MiniCPM-V 4.7 1B vs Ripple-65B Preview
- MiniCPM-V 4.7 1B vs SmolLM3-3B
- MiniCPM-V 4.7 1B vs Spark-X2.6-4B
- MiniCPM-V 4.7 1B vs Swift 1.5 Qwen3.8-27B
- MiniCPM-V 4.7 1B vs Swift 1.5 Qwen3.8-Flash-Next
- MiniCPM-V 4.7 1B vs ThinkingCap (Qwen3.8 27B)
- MiniCPM-V 4.7 1B vs Baidu Unlimited-OCR-2
Machine-readable
- This page as Markdown
/compare/gemma-4-e2b-it-vs-minicpm-v-4-7-1b.md· slim.min.md· JSON.json(or sendAccept: text/markdown) - Each listing in full
/api/v1/tools/gemma-4-e2b-it.json·/api/v1/tools/minicpm-v-4-7-1b.json - From a terminal
anchor compare gemma-4-e2b-it minicpm-v-4-7-1b(the CLI) - Over MCP
compare_tools {"a": "gemma-4-e2b-it", "b": "minicpm-v-4-7-1b"}at/mcp, no key