The story so far
Weights, every dated article, and repo commits on one rail.
12 repo commits since the weights landed
Teal is the weights (2024-01-25), grey ticks are repo commits.
deepseek-coder-7b-instruct-v1.5
deepseek-ai DeepSeek License (deepseek) out 3 years ago
6.91B parameters, text in, text out, 4K tokens of context.
Verified 2026-10-25 Rank #463 in LLMs · score 77.6 method v1 Sources 3 Downloads 30d 416.6k · #180 of 619 tracked
Do people use it?416.6kdownloads, 30 days#180 of 619 we track · top 29%+3 likes, downloads −19% this week191 likesHugging Face2026-10-24
Where it's used
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 13 quantized builds and 31 Spaces.
- As a service0no API provider listed
- On own hardware13quantized builds
- Built on12151 fine-tunes · 31 adapters · 39 merges
- In apps31Hugging Face Spaces
Top Spaces running it: KBaba7/Quant ♥ 51 · FallnAI/Quantize-HF-Models ♥ 19 · bhaskartripathi/LLM_Quantization ♥ 16 · openfree/LLM_Quantization ♥ 16
Will it fit?
Weights only, an estimate: leave room for the KV cache. Method
Fits a 16 GB card at INT4
About 4.1 GB at INT4, 8.3 GB at FP8, 17 GB at BF16: weights only, so leave room for the KV cache.
Estimate params × bytes per param × 1.2 · how we estimate
What people built on it
134 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
134 derivatives on the Hub, led by 51 fine-tunes
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Reference
Everything else, folded: the full tables, every source.
What the card says
1 cited sentence
Use-case evidence
3 use cases
What it is good at 3 use cases
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
Facts
19 fields
Listed in LLMs
Facts
- deepseek-coder-7b-instruct-v1.5 has 6.91B parameters. (safetensors metadata, observed 2026-10-24)
- Context length is 4,096 tokens (4K). (config.json max_position_embeddings, observed 2026-10-24)
- License: DeepSeek License (deepseek) (friction: attribution). (license file, observed 2026-10-24)
- Created 2024-01-25, last modified 2024-02-05. (HF API, observed 2026-10-24)
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | LlamaForCausalLM |
| Model type | llama |
| Dtype | bfloat16 |
| Params by dtype | BF16: 6,910,365,696 |
| Size band | M |
| MoE | no |
| Context length | 4,096 (config.json max_position_embeddings) |
| License | DeepSeek License (deepseek) (other) |
| Hub snapshot | 2026-10-25 |
Servability
vLLM · SGLang
Servability
vLLM: supported · LlamaForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · LlamaForCausalLM · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Hardware fit
3 estimates
Hardware fit ESTIMATE
Weights-only memory for deepseek-coder-7b-instruct-v1.5: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 17 GB · Single GPU (<24 GB) |
|---|---|
| FP8 (1 byte/param) | 8.3 GB · Single GPU (<24 GB) |
| INT4 (0.5 bytes/param) | 4.1 GB · Single GPU (<24 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra.
All variants
40 variants
Lineage
quantized: 13finetune: 51adapter: 31merge: 39
Top variants (40 shown, ranked by downloads)
Repository changelog
12 commits
Repository changelog 12 commits
| Date | Commit | SHA |
|---|---|---|
| 2024-02-05 | Update generation_config.json | 2a050a4 |
| 2024-02-05 | Update config.json | c63ca91 |
| 2024-02-05 | Update generation_config.json | 0c5229e |
| 2024-02-04 | Update README.md | f701335 |
| 2024-02-04 | Update README.md | 3dd5d4e |
| 2024-01-26 | Update README.md | 8a11174 |
| 2024-01-26 | Update README.md | c4ba587 |
| 2024-01-26 | Update README.md | 70d4fed |
| 2024-01-26 | Update README.md | 194ff74 |
| 2024-01-26 | Update README.md | 053b477 |
| 2024-01-25 | upload models | e66d398 |
| 2024-01-25 | initial commit | 46a1833 |
Similar models
3 models
Similar models
- Llama-3.1-8B-Instruct · 8.03B same size band (M), same task (text-generation), more providers
- Qwen2.5-7B-Instruct · 7.62B same size band (M), same task (text-generation), more providers, permissive license
- Mistral-7B-Instruct-v0.2 · 7.24B same size band (M), same task (text-generation), more providers, permissive license
Sources
3 sources
Sources
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:39 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:10 UTC · Last verified 2026-10-25
Follow deepseek-coder-7b-instruct-v1.5
Watch this model
- /feeds/models/deepseek-ai--deepseek-coder-7b-instruct-v1.5.xml(RSS 2.0) · every change to deepseek-coder-7b-instruct-v1.5
- /feeds/families/deepseek-ai--deepseek-v1.5.xml(RSS 2.0) · every change across the deepseek-v1.5 family
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.