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: 1,048,576 tokens.
Max completion tokens on that row: 262,144.
Input / output on that row: $2.00 / $6.00 per 1M tokens.
Architecture fields: text->text · Qwen.
Hugging Face id on that row: Qwen/Qwen3.8-2.4T-A95B.
Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of [Qwen3.8 Max](/qwen/qwen3.8-max), with 95 billion active parameters out of 2.4 trillion total.
When you want the 2026-08-12 catalog SKU, not a later rename.
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
We did not open a vals.ai card for this identity. Chip omitted.
Ranked view: Best coding models: the independent leaderboard puts this row and every other card that carries an independent coding score on one board, with the confidence intervals left visible.
Copied from AA or vals HTML we opened. 0.0% placeholder bars are omitted. Do not average these into the hero tiles, and do not invent CI, $/task, or steps for Artificial Analysis.
| Bench | Printed | Note | Source | URL | As of |
|---|---|---|---|---|---|
| GDPval-AA v2 | 61% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| τ³-Banking | 49.1% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| Terminal-Bench v2.1 | 82% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| SciCode | 51.6% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| Humanity's Last Exam | 42.4% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| GPQA Diamond | 93.5% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| CritPt | 20% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| AA-Omniscience Accuracy | 31.3% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
| AA-LCR | 75.3% | Printed on the AA Intelligence Evaluations grid for Qwen3.8 2.4T A95B. Not a DeepSWE chip. | Artificial Analysis Qwen3.8 2.4T A95B | source | 2026-08-19 |
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: Qwen/Qwen3.8-2.4T-A95B. 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 |
|---|---|
model-00001-of-00213.safetensors | 34,359,738,528 |
model-00002-of-00213.safetensors | 17,179,869,336 |
model-00003-of-00213.safetensors | 34,359,738,528 |
model-00004-of-00213.safetensors | 17,179,869,336 |
model-00005-of-00213.safetensors | 34,359,738,528 |
model-00006-of-00213.safetensors | 17,179,869,336 |
model-00007-of-00213.safetensors | 34,359,738,528 |
model-00008-of-00213.safetensors | 17,179,869,336 |
model-00009-of-00213.safetensors | 34,359,738,528 |
model-00010-of-00213.safetensors | 17,179,869,336 |
model-00011-of-00213.safetensors | 3,985,468,128 |
model-00012-of-00213.safetensors | 34,359,738,528 |
model-00013-of-00213.safetensors | 17,179,869,336 |
model-00014-of-00213.safetensors | 34,359,738,528 |
model-00015-of-00213.safetensors | 17,179,869,336 |
model-00016-of-00213.safetensors | 34,359,738,528 |
model-00017-of-00213.safetensors | 17,179,869,336 |
model-00018-of-00213.safetensors | 34,359,738,528 |
model-00019-of-00213.safetensors | 17,179,869,336 |
model-00020-of-00213.safetensors | 3,972,721,928 |
model-00021-of-00213.safetensors | 34,359,738,528 |
model-00022-of-00213.safetensors | 17,179,869,336 |
model-00023-of-00213.safetensors | 34,359,738,528 |
model-00024-of-00213.safetensors | 17,179,869,336 |
model-00025-of-00213.safetensors | 4,068,475,024 |
model-00026-of-00213.safetensors | 2,810,632,576 |
model-00027-of-00213.safetensors | 34,359,738,528 |
model-00028-of-00213.safetensors | 17,179,869,336 |
model-00029-of-00213.safetensors | 34,359,738,528 |
model-00030-of-00213.safetensors | 17,179,869,336 |
model-00031-of-00213.safetensors | 34,359,738,528 |
model-00032-of-00213.safetensors | 17,179,869,336 |
model-00033-of-00213.safetensors | 34,359,738,528 |
model-00034-of-00213.safetensors | 17,179,869,336 |
model-00035-of-00213.safetensors | 34,359,738,528 |
model-00036-of-00213.safetensors | 17,179,869,336 |
model-00037-of-00213.safetensors | 3,981,110,080 |
model-00038-of-00213.safetensors | 34,359,738,528 |
model-00039-of-00213.safetensors | 17,179,869,336 |
model-00040-of-00213.safetensors | 34,359,738,528 |
model-00041-of-00213.safetensors | 17,179,869,336 |
model-00042-of-00213.safetensors | 34,359,738,528 |
model-00043-of-00213.safetensors | 17,179,869,336 |
model-00044-of-00213.safetensors | 34,359,738,528 |
model-00045-of-00213.safetensors | 17,179,869,336 |
model-00046-of-00213.safetensors | 3,939,167,440 |
model-00047-of-00213.safetensors | 34,359,738,528 |
model-00048-of-00213.safetensors | 17,179,869,336 |
model-00049-of-00213.safetensors | 34,359,738,528 |
model-00050-of-00213.safetensors | 17,179,869,336 |
model-00051-of-00213.safetensors | 2,810,632,616 |
model-00052-of-00213.safetensors | 34,359,738,528 |
model-00053-of-00213.safetensors | 17,179,869,336 |
model-00054-of-00213.safetensors | 34,359,738,528 |
model-00055-of-00213.safetensors | 17,179,869,336 |
model-00056-of-00213.safetensors | 34,359,738,528 |
model-00057-of-00213.safetensors | 17,179,869,336 |
model-00058-of-00213.safetensors | 34,359,738,528 |
model-00059-of-00213.safetensors | 17,179,869,336 |
model-00060-of-00213.safetensors | 3,909,971,448 |
model-00061-of-00213.safetensors | 34,359,738,528 |
model-00062-of-00213.safetensors | 17,179,869,336 |
model-00063-of-00213.safetensors | 34,359,738,528 |
model-00064-of-00213.safetensors | 17,179,869,336 |
tokenizer.json | 12,809,320 |
vocab.json | 6,722,759 |
merges.txt | 3,353,259 |
model.safetensors.index.json | 137,777 |
README.md | 35,755 |
tokenizer_config.json | 16,438 |
chat_template.jinja | 7,495 |
config.json | 3,951 |
LICENSE | 3,390 |
.gitattributes | 1,570 |
generation_config.json | 202 |
Showing the first 64 of 213 safetensor shards, plus 11 other files. The snapshot was capped at 64 shards when the tree API was read on 2026-08-19, so the remaining 149 are on the Files tab, not on this page. Totals above are the full tree API sum, not this slice. Byte sizes are Hub tree API size fields, never invented.
OpenRouter slug qwen/qwen3.8-2.4t-a95b on the 2026-08-19 catalog.
First party: https://qwen.ai/.
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: $2.00 in / $6.00 out per 1M tokens. Context 1,048,576 in / 262,144 out.
No separate internal-reasoning price on that row.
| Claim | Source | As of |
|---|---|---|
| OpenRouter id qwen/qwen3.8-2.4t-a95b; context 1048576; created 1786551702; $2.00 / $6.00 per 1M | OpenRouter /api/v1/models | 2026-08-19 |
| AA Intelligence Index 58; 9 printed Index benches | Artificial Analysis model page | 2026-08-19 |
| OpenRouter id qwen/qwen3.8-2.4t-a95b, context 1,048,576 | OpenRouter /api/v1/models | 2026-08-19 |
| HF downloads 12,699 | Hugging Face API Qwen/Qwen3.8-2.4T-A95B | 2026-08-19 |
| HF safetensors.total: 2,446,182,725,504 stored tensors; HF API usedStorage 4,892,378,458,656 bytes; created 2026-08-08T01:50:52.000Z; HF tensors BF16 | Hugging Face API Qwen/Qwen3.8-2.4T-A95B | 2026-08-19 |
| 213 safetensor shards; shard bytes 4,892,365,649,336; tree files 224 | Hugging Face tree API Qwen/Qwen3.8-2.4T-A95B | 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 Qwen's model family with Qwen3.7 Flash.
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 qwen/qwen3.8-2.4t-a95b.