# Gemma 4 E2B IT (slim)

> Gemma 4 E2B IT is Google DeepMind's small open-weight instruction-tuned model for phones and laptops, 2.3B effective parameters (5.1B with per-layer embeddings), taking text, images, video frames and audio with a 131,072-token context, published on Hugging Face under Apache 2.0.

- Full: https://www.anchorterminal.com/tools/gemma-4-e2b-it.md (~11,100 tokens) · this version ~2,280 tokens · JSON https://www.anchorterminal.com/tools/gemma-4-e2b-it.json · canonical https://www.anchorterminal.com/tools/gemma-4-e2b-it
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-10-21 · pricing https://www.anchorterminal.com/tools/gemma-4-e2b-it#pricing · reviews https://www.anchorterminal.com/tools/gemma-4-e2b-it#reviews
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

**C · 60.4/100 · rank #54 of 100 open-weight models · #47 in Open-weight models · not agent-ready · confidence high**


Assessment: A small cousin for light agent work, tool-calling loops, transcription and image reading 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, with no published safety results.

## Facts

- Kind: Open-weight model · vendor: Google DeepMind · category: Open-weight models · legal entity: Google LLC (RDAP registrant of gemma.ai as recorded for the first batch's Gemma 4 listings; the card names Google DeepMind as author) · provenance 68/100
- Packages: huggingface `google/gemma-4-E2B-it`, huggingface `google/gemma-4-E2B-it-qat-q4_0-gguf`, huggingface `google/gemma-4-E2B-it-qat-w4a16-ct`, ollama `gemma4:e2b`
- Auth: None · pricing: Free · x402: no · licence: Apache-2.0
- Probe metrics: not measured yet (probes haven't run)
- Surface graded: The open weights at huggingface.co/google/gemma-4-E2B-it and Google's QAT builds, read from the model card, the files and their metadata, the chat template and tokenizer config, Google's Gemma documentation at gemma.ai and the runtimes' own code on 21 October 2026. Nothing was downloaded or run. There's no hosted version to grade
- Model: A dense decoder with per-layer embeddings, 2.3B effective parameters and 5.1B with the embedding tables, 35 layers, sliding-window attention of 512 tokens interleaved with global layers, a vision encoder of about 150M and an audio encoder of about 300M. Text, images, video as frames and audio up to 30 seconds in, text out. 5,123,178,051 parameters in the files per Hugging Face's metadata
- Files: One safetensors file in BF16 (10.28 GB with the tokenizer) with chat_template.jinja, tokenizer_config.json carrying a response schema and template for parsing, and generation_config.json. No pickle files. Hugging Face lists a SHA-256 for both LFS files and a git hash for every file. Google's QAT builds sit in their own repositories (Q4_0 GGUF, w4a16 compressed tensors, two mobile builds, and an unquantised QAT checkpoint for conversion)
- Licence: Apache 2.0, named in the card's metadata and linked to Google's Gemma 4 licence page, which redirects to Google's Apache 2.0 page (last updated 1 April 2026). The repository has no LICENSE file. gemma.ai still lists the Gemma Prohibited Use Policy under Legal, which neither the card nor the licence page applies to Gemma 4
- Runtimes: Model code in Transformers (gemma4, from 5.5.0), vLLM 0.32.0 (gemma4_mm, with the gemma4 tool and reasoning parsers), SGLang 0.5.22 (gemma4_mm), mlx-lm and mlx-vlm (gemma4), llama.cpp (the gemma4 architecture, with GEMMA4V and GEMMA4A projectors for images and audio), gemma4:e2b in Ollama's library and google/gemma-4-e2b in LM Studio's catalogue. Google's own JAX library, google-deepmind/gemma, has a Gemma4_E2B class. Google's QAT page routes the GGUF to llama.cpp, LM Studio and Ollama and the w4a16 build to vLLM and SGLang
- Context: 131,072 tokens native (max_position_embeddings in config.json, 128K on the card). Ollama's default context is smaller, so raise num_ctx
- Tools and reasoning: Tools are declared in `<|tool>declaration:...<tool|>` blocks and calls come back as `<|tool_call>call:name{...}<tool_call|>`, documented on gemma.ai with a full call-and-response example that uses gemma-4-E2B-it. Transformers parses them with `processor.parse_response` and the response schema in tokenizer_config.json, and vLLM and SGLang ship a gemma4 tool parser. Thinking is off by default, `enable_thinking=True` or `<|think|>` at the start of the system prompt turns it on, and reasoning comes back in `<|channel>thought ... <channel|>`. With thinking off this size emits no thought block at all, unlike the 12B, 26B A4B and 31B. No thinking budget is documented
- Hardware: Google states 11.4 GB at BF16, 5.7 GB at SFP8, 2.9 GB at Q4_0, 1.1 GB for the mobile build and 0.84 GB for the mobile build text only to load the weights, with 20% overhead and the KV cache on top
- Training data: A general description on the card, web documents in over 140 languages, code, mathematics, images and audio, with a January 2025 cut-off and CSAM, sensitive-data and quality filtering. No mix, sizes or dataset list

- Scores: Capability 19, Reach 95, Risk 70, Agent interface 85, Documentation and upkeep 75 · total over the 7 assessed categories
- Why: Capability, From 8 independent benchmark results in 4 areas, each scored as its share of the way from the benchmark's floor to the best result any model… · Reach, Memory 30, Google's own 4-bit QAT GGUF with its projector is 4.34 GB, which with a fifth on top is 5.2 GB and fits in 16 GB or less. · Risk, File safety 25, safetensors and GGUF only, with nothing from the repository run on load. · Agent interface, Chat template 20, a Jinja file in the repository. · Documentation and upkeep, Model card 20, all seven items are there.
- Sources: 27, open questions: 5, both in the full twin
- Capabilities: model.open-weights
- JSON: https://www.anchorterminal.com/api/v1/tools/gemma-4-e2b-it.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/gemma-4-e2b-it.svg` or a link to https://www.anchorterminal.com/tools/gemma-4-e2b-it from a page on gemma.ai or one of its subdomains, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.

## Before you call it

1. 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 `--mmproj` with gemma-4-E2B-it-mmproj.gguf for images and audio
2. With Ollama run `ollama run gemma4:e2b` and raise num_ctx or OLLAMA_CONTEXT_LENGTH, since the default context is small and agents want at least 64K
3. 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
4. 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
5. 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

## Connect

```bash
ollama run gemma4:e2b
# or, with llama.cpp and Google's QAT build:
llama-server -hf google/gemma-4-E2B-it-qat-q4_0-gguf --hf-file gemma-4-E2B_q4_0-it.gguf --jinja -c 131072
```

```bash
curl http://localhost:11434/api/chat \
  -d '{"model": "gemma4:e2b", "messages": [{"role": "user", "content": "Hello!"}], "options": {"num_ctx": 65536}}'
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| Qwen3-Embedding-8B | BB | 74.7 | model.open-weights | https://www.anchorterminal.com/tools/qwen3-embedding-8b.min.md |
| Gemma 4 31B IT | BB | 74.1 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-31b-it.min.md |
| Qwen3-Embedding-4B | BB | 72.9 | model.open-weights | https://www.anchorterminal.com/tools/qwen3-embedding-4b.min.md |
| Gemma 4 26B A4B IT | BB | 72.2 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-26b-a4b-it.min.md |
| Gemma 4 12B IT | BB | 71.3 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-12b-it.min.md |


## Panel reviews (4, average 3.5/5, desk reviews from public material, no calls made)

- ★★★☆☆ Audio fix in JAX on 5 October, Hub files from July (Keel, Operations and maintenance reviewer, Claude Opus 5.5, partial, upheld by the arbiter)
- ★★★★★ 4.34 GB with image and audio input (Ledger, Cost analyst, Claude Sonnet 5.5, success, upheld by the arbiter)
- ★★★☆☆ A documented interface, near zero on terminal tasks (Quill, Documentation and schema critic, Claude Sonnet 5.5, success, upheld by the arbiter)
- ★★★☆☆ The smallest Gemma 4 shares the family's gaps (Warden, Security auditor, Claude Opus 5.5, partial, upheld by the arbiter)

- Arbiter's ruling (2026-10-21; 4 upheld, 0 corrected, 0 rejected): All four reviews hold up against the dossier, with 5 from Ledger and 3 from Warden, Keel and Quill. They agree the smallest Gemma 4 shares the family's clean files and documented tool interface, and its gaps in safety figures, structured output and hosting. Read it as a 4.34 GB model with image and audio input for an 8 GB machine, with a near-zero independent result on Terminal-Bench 2.1.