DeepSWE
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
Apache-2.0 on the official repo. Not listed as a Continuum hosted id. Catalog created 6 Jul 2026 (unix 1783344048). 262,144 in / 128,000 out; text→text. Hugging Face id tencent/Hy3.
HF config.json (fetched 19 Aug 2026): HYV3ForCausalLM, 80 layers, hidden 4096, 192 experts, 8 experts/token, max position 262,144. Chat template is present. No arXiv tag on the Hub card we fetched.
Week-token rank is #2 on OpenRouter This Week (9.83T). That is volume, not a coding score, and it is why this card exists next to models with DeepSWE rows.
Reach for Hy3 when you want the open Tencent checkpoint that actually moved tokens this week. AA lists Intelligence Index 42. We do not have an official DeepSWE or vals card.
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 |
|---|---|---|---|---|---|
| AA list pair | $0.14 / $0.55 per 1M | Printed AA pricing sentence. OpenRouter catalog is $0.132 / $0.528 plus $0.033/M cache read. | Artificial Analysis Hy3 | source | 2026-08-19 |
| AA throughput | 70.0 tok/s | Printed Speed row (“At 70 tokens per second”). Not a coding score. | Artificial Analysis Hy3 | source | 2026-08-19 |
| AA Index eval cost | $93.20 | Printed “it cost $93.20 to evaluate Hy3 on the Intelligence Index.” Not $/task on DeepSWE. | Artificial Analysis Hy3 | 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.
Community quants: Community quants are community.
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-00099.safetensors | 7,247,829,008 |
model-00002-of-00099.safetensors | 7,247,829,584 |
model-00003-of-00099.safetensors | 7,247,829,584 |
model-00004-of-00099.safetensors | 7,247,829,584 |
model-00005-of-00099.safetensors | 7,247,829,584 |
model-00006-of-00099.safetensors | 1,073,745,312 |
model-00007-of-00099.safetensors | 7,247,829,008 |
model-00008-of-00099.safetensors | 7,247,829,584 |
model-00009-of-00099.safetensors | 7,247,829,584 |
model-00010-of-00099.safetensors | 7,247,829,584 |
model-00011-of-00099.safetensors | 7,247,829,584 |
model-00012-of-00099.safetensors | 1,073,742,272 |
Showing the first 12 of 99 safetensor shards. The tree also holds 45 non-shard files, none of them listed here. The snapshot was capped at 12 shards when the tree API was read on 2026-08-19, so the remaining 87 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.
Not listed as a Continuum hosted id. First-party and OpenRouter tencent/hy3.
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 catalog 19 Aug 2026: $0.132 / $0.528 per million, cache read $0.033 /M, created 1783344048. No DeepSWE $/task.
Artificial Analysis Hy3 page (19 Aug 2026) prints $0.14 / $0.55 per 1M, 70.0 tok/s, and “it cost $93.20 to evaluate Hy3 on the Intelligence Index.” Cite both; do not average them into OpenRouter.
Artificial Analysis Intelligence Index 42, fetched 19 Aug 2026. Same page prints $0.14 / $0.55, 70.0 tok/s, and $93.20 to run the Index. The AA tile is the published integer 42. AA did not print a DeepSWE-style CI, step count, or per-eval percent in the server HTML we saved. Those nine Index eval names stay unnamed here.
| Claim | Source | As of |
|---|---|---|
| OR catalog $0.132 / $0.528, cache read $0.033/M, 262,144 / 128,000, created 1783344048, hugging_face_id tencent/Hy3 | OpenRouter catalog | 2026-08-19 |
| 9.83T week tokens, rank 2 | OpenRouter rankings This Week | 2026-08-19 |
| HF safetensors.total 298,786,155,776; usedStorage 597,584,438,620; 99 shards; likes 926; created 2026-07-02T13:23:54Z; Apache-2.0 | Hugging Face API tencent/Hy3 | 2026-08-19 |
| AA Index 42; printed $0.14 / $0.55; 70.0 tok/s; $93.20 to evaluate Index | Artificial Analysis Hy3 | 2026-08-19 |
| vals.ai tencent_hy3 404 | vals.ai Hy3 | 2026-08-19 |
| OpenRouter id tencent/hy3, context 262,144 | OpenRouter /api/v1/models | 2026-08-19 |
| HF downloads 12,070 | Hugging Face API tencent/Hy3 | 2026-08-19 |
| 99 safetensor shards; shard bytes 597,578,239,288; tree files 144 | Hugging Face tree API tencent/Hy3 | 2026-08-19 |
| Spaces API returned 22 | Hugging Face Spaces / collections API | 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 Tencent's model family with Tencent HY3 Preview.
The official DeepSWE board updated 13 Aug 2026 did not list Hy3. Omitting is the rule.
https://www.vals.ai/models/tencent_hy3 returned 404 on 19 Aug 2026. Chip omitted.
AA printed $0.14 / $0.55. OpenRouter printed $0.132 / $0.528 plus $0.033/M cache. We cite both.
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 tencent/hy3.