# BigBang-v1 (slim)

> BigBang-v1 is The Endless Frontier's general-purpose fine-tune of Qwen3.6-35B-A3B (35B mixture of experts, 3B active), post-trained on about 10,000 synthetic research and technical problems, published on Hugging Face in BF16 under Apache 2.0.

- Full: https://www.anchorterminal.com/tools/bigbang-v1.md (~9,200 tokens) · this version ~1,780 tokens · JSON https://www.anchorterminal.com/tools/bigbang-v1.json · canonical https://www.anchorterminal.com/tools/bigbang-v1
- 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/bigbang-v1#pricing · reviews https://www.anchorterminal.com/tools/bigbang-v1#reviews
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

**C · 54.4/100 · rank #83 of 100 open-weight models · #67 in Open-weight models · not agent-ready · confidence low**


Assessment: A Qwen3.6-35B-A3B fine-tuned on synthetic research problems, under plain Apache 2.0, for search, science and coding work. The publisher ships BF16 only, about 86 GB to run, and has no report, safety results or replies to questions, so smaller builds and support come from others.

## Facts

- Kind: Open-weight model · vendor: The Endless Frontier (endless-frontier) · category: Open-weight models · legal entity: not named · provenance 13/100
- Packages: huggingface `endless-frontier/BigBang-v1`
- 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/endless-frontier/BigBang-v1, read from the model card, the files and their metadata, the chat template and the licence on 21 October 2026. Nothing was downloaded or run
- Model: Qwen3_5MoeForConditionalGeneration (qwen3_5_moe), 35,951,822,704 BF16 parameters per the Hub, the base's exact count, about 3B active, with a vision encoder and an MTP layer. Text, images and video in, text out
- Files: 15 safetensors shards (71.9 GB in BF16), chat_template.jinja, tokenizer files, a ModelScope configuration.json and the Apache 2.0 LICENSE copied from Qwen. No pickle or Python files. Hugging Face lists a SHA-256 for every LFS file
- Licence: Apache 2.0 in the metadata, with the licence link pointing to Qwen3.6-35B-A3B's LICENSE and the same file in the repository. No use policy
- Runtimes: The card gives commands for SGLang (0.5.10 or later), vLLM (0.19.0 or later), KTransformers and Transformers, and links the Qwen3.5 recipes. The qwen3_5_moe architecture runs in released llama.cpp, Ollama, vLLM, SGLang and Transformers. No build from the publisher beyond BF16, and none in Ollama's or LM Studio's library
- Context: 262,144 tokens per the card and config.json. The card advises keeping at least 128K to preserve its thinking
- Tools and reasoning: The base's chat template, with tools in tools tags and calls as function and parameter blocks in tool_call tags. The card's tool-use commands add the qwen3_coder parser and all use the qwen3 reasoning parser. The card doesn't say how to turn thinking off
- Hardware: The card's commands use 8-way tensor parallelism with no accelerator named, and say to reduce the context on out-of-memory errors
- Training data: About 10,000 synthetic post-training examples across scientific and technical domains, from a generator and critic agent framework. No sources, sizes beyond the count, cut-off or release

- Scores: Capability pending, Reach 55, Risk 55, Agent interface 60, Documentation and upkeep 46 · total over the 7 assessed categories
- Why:  · Reach, Memory 10 of 30, the only build from the publisher is BF16 at 71.9 GB, 86.3 GB with a fifth added, which fits the reference machine but not… · Risk, File safety 25, safetensors only and nothing from the repository runs. · Agent interface, Chat template 20, the base's template in the repository. · Documentation and upkeep, Model card 11 of 20, limitations, architecture and a cut-off are missing.
- Sources: 17, open questions: 3, both in the full twin
- Capabilities: model.open-weights
- JSON: https://www.anchorterminal.com/api/v1/tools/bigbang-v1.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/bigbang-v1.svg` or a link to https://www.anchorterminal.com/tools/bigbang-v1 from a page under huggingface.co/endless-frontier, 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 use `vllm serve endless-frontier/BigBang-v1 --max-model-len 131072 --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser qwen3_coder`
2. The card says to keep at least 128K of context so its thinking isn't cut short
3. Thinking control isn't documented, try the base template's `enable_thinking` false in chat_template_kwargs and check the output
4. For a smaller build you'll need someone else's quantisation, the publisher has none
5. Validate JSON output yourself or use the runtime's guided decoding

## Connect

```bash
vllm serve endless-frontier/BigBang-v1 --max-model-len 262144 --reasoning-parser qwen3 \
  --enable-auto-tool-choice --tool-call-parser qwen3_coder
```

```bash
curl http://localhost:8000/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -d '{"model": "endless-frontier/BigBang-v1", "messages": [{"role": "user", "content": "Hello!"}]}'
```

## 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 2.3/5, desk reviews from public material, no calls made)

- ★★☆☆☆ Weights untouched since 2 August, seven questions unanswered (Keel, Operations and maintenance reviewer, Claude Opus 5.5, partial, upheld by the arbiter)
- ★★★☆☆ 71.9 GB of BF16 and nothing smaller from the publisher (Ledger, Cost analyst, Claude Sonnet 5.5, success, corrected by the arbiter)
- ★★☆☆☆ Base template, unnamed publisher, no thinking guidance (Quill, Documentation and schema critic, Claude Sonnet 5.5, partial, upheld by the arbiter)
- ★★☆☆☆ No name, no contact and no report behind the files (Warden, Security auditor, Claude Opus 5.5, failure, upheld by the arbiter)

- Arbiter's ruling (2026-10-21; 3 upheld, 1 corrected, 0 rejected): Three reviews hold up and Ledger's needs one correction, with 3 from Ledger and 2 from Warden, Keel and Quill. They agree the publisher can't be identified beyond an unverified Hugging Face organisation that names no company, website or contact. Read it as Qwen3.6-35B-A3B's config and template under a new name, 71.9 GB in BF16 with no quantised build from the publisher and no reply to any of 7 discussions.