DeepSWE
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
Context window on the 2026-08-19 OpenRouter row: 262,144 tokens.
Max completion tokens were not listed on that row.
Input / output on that row: $0.20 / $0.20 per 1M tokens.
Architecture fields: text+image->text · Mistral.
Hugging Face id on that row: mistralai/Ministral-3-14B-Instruct-2512.
The largest model in the Ministral 3 family, Ministral 3 14B offers frontier capabilities and performance comparable to its larger Mistral Small 3.2 24B counterpart.
When you want the 2025-12-02 catalog SKU, not a later rename.
No independent board lists this exact SKU as of 2026-08-19.
No other card from this lab carries an independent board row either, so there is no sibling worth pointing at.
Boards checked: DeepSWE official (2026-08-13), Artificial Analysis Intelligence Index, vals.ai hero index. Every card that does carry a row is on the coding leaderboard.
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
No Intelligence Index integer fetched for this identity. Chip omitted.
We did not open a vals.ai card for this identity. Chip omitted.
Plus is $25/mo with $25 weekly hosted usage. The Mac app stays free with your own keys. Get Plus is not Download for Mac.
Official weights id on the 2026-08-19 OpenRouter row: mistralai/Ministral-3-14B-Instruct-2512. Community GGUF is a quant, not a second Continuum card.
This is the model repo, not Download for Mac. Downloads and likes from the Hugging Face API on 2026-08-19.
usedStorage and the sum of .safetensors sizes can differ (LFS pointers, non-shard files, duplicate copies). Cited separately. Source: recursive tree API, 2026-08-19.
| Path | Bytes |
|---|---|
consolidated.safetensors | 15,726,706,472 |
model-00001-of-00004.safetensors | 4,980,912,374 |
model-00002-of-00004.safetensors | 4,996,851,408 |
model-00003-of-00004.safetensors | 4,993,872,954 |
model-00004-of-00004.safetensors | 755,093,832 |
tokenizer.json | 17,077,420 |
tekken.json | 16,753,784 |
model.safetensors.index.json | 117,404 |
tokenizer_config.json | 21,177 |
README.md | 20,322 |
chat_template.jinja | 11,913 |
SYSTEM_PROMPT.txt | 2,362 |
config.json | 1,906 |
.gitattributes | 1,618 |
params.json | 1,188 |
processor_config.json | 976 |
generation_config.json | 131 |
All 17 files the tree API returned are listed: every one of the 5 safetensor shards plus 12 other files. Files tab: https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512/tree/main.
OpenRouter slug mistralai/ministral-14b-2512 on the 2026-08-19 catalog.
First party: https://docs.mistral.ai/getting-started/models/.
Not on the Continuum host list fetched 2026-08-19.
We list first-party and OpenRouter, plus Continuum only when the live public allowlist named the id. This is not a 15-host routing table.
OpenRouter 2026-08-19: $0.20 in / $0.20 out per 1M tokens. Context 262,144 in.
No separate internal-reasoning price on that row.
| Claim | Source | As of |
|---|---|---|
| OpenRouter id mistralai/ministral-14b-2512; context 262144; created 1764681735; $0.20 / $0.20 per 1M | OpenRouter /api/v1/models | 2026-08-19 |
| OpenRouter id mistralai/ministral-14b-2512, context 262,144 | OpenRouter /api/v1/models | 2026-08-19 |
| HF downloads 215,659 | Hugging Face API mistralai/Ministral-3-14B-Instruct-2512 | 2026-08-19 |
| HF safetensors.total: 13,945,032,240 stored tensors; HF API usedStorage 31,521,099,423 bytes; created 2025-10-31T08:43:24.000Z; HF tensors BF16 + F8_E4M3 | Hugging Face API mistralai/Ministral-3-14B-Instruct-2512 | 2026-08-19 |
| 5 safetensor shards; shard bytes 31,453,437,040; tree files 17 | Hugging Face tree API mistralai/Ministral-3-14B-Instruct-2512 | 2026-08-19 |
No other card in this cluster shares a published DeepSWE official row we can put next to this one. We will not compare on vendor-blog numbers.
Continue through Mistral's model family with Ministral 3 8B 2512.
Lab posts pick a harness, an effort, a split, and sometimes a private eval set. DeepSWE publishes the official mini-swe-agent row with cost, tokens, and steps. Scale SWE-bench Pro only counts when the public shared-harness board has a row we can fetch. A higher lab number is usually a different test, not a better one.
No Continuum hosted id is verified for this model, so this page does not invent a curl target. Use the first-party API or the OpenRouter slug mistralai/ministral-14b-2512.