DeepSWE
No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.
Buyers search the name. This is a grounded search API, not a SWE harness winner. OpenRouter does not even expose tools on this slug. Official Perplexity pricing.md prints $3 / $15 per 1M plus a per-request fee. AA exists but prints no copyable Intelligence Index integer.
Closed search SKU. OpenRouter catalog 19 Aug 2026: perplexity/sonar-pro, 200,000 in / 8,000 out; text + image in, text out. hugging_face_id empty. Created unix 1741312423 (7 Mar 2025). Not listed as a Continuum hosted id.
OpenRouter supported_parameters for this slug: frequency_penalty, max_tokens, presence_penalty, temperature, top_k, top_p, web_search_options. Tools are absent. That is the product: cited web answers, not a repo agent.
Use Sonar Pro when the job is a cited web answer with live retrieval. Official Perplexity pricing.md prints token rates plus a per-request fee. OpenRouter’s catalog note says the listed $3 / $15 already includes Perplexity search pricing and also prints web_search $0.005. Cite both pages; do not invent a third fee.
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.
vals.ai /models/perplexity_sonar-pro returned 404 on 19 Aug 2026. Chip omitted.
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 |
|---|---|---|---|---|---|
| Official token pair | $3 / $15 per 1M | Printed on Perplexity pricing.md for sonar-pro. Matches the OpenRouter list pair. | Perplexity pricing.md | source | 2026-08-19 |
| Official request fee | $6 / $10 / $14 per 1,000 requests | Printed low / medium / high on the same official page. A second line, not a token rate. | Perplexity pricing.md | source | 2026-08-19 |
| OpenRouter web_search field | $0.005 | Catalog field next to $3 / $15. Catalog note says token rates already include search pricing. | OpenRouter catalog | source | 2026-08-19 |
| AA printed pair | $0.00 / $0.00 per 1M | Printed In/Out on the AA Sonar Pro page. Not the official invoice. Do not chip it. | Artificial Analysis Sonar Pro | 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.
Perplexity does not publish Sonar weights. No HF button.
Lab card and OpenRouter listing only. No Hugging Face Files button, because there is no official repo to point at.
Fetched OpenRouter catalog on 2026-08-19. hugging_face_id was empty.
Not listed as a Continuum hosted id. First-party Sonar API or OpenRouter perplexity/sonar-pro.
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.
Official Perplexity pricing.md, 19 Aug 2026: $3 / $15 per million plus $6 / $10 / $14 per 1,000 requests (low / medium / high). OpenRouter catalog prints $3 / $15 and web_search $0.005 and says token rates include search pricing. Cite both pages; do not invent a third fee.
No DeepSWE $/task. This slug has no tools on the catalog. AA’s printed pair on the same day is $0.00 / $0.00, not the invoice.
Artificial Analysis Sonar Pro page (200, 19 Aug 2026) prints In $0.00 / Out $0.00, “Reasoning No”, and the screen-reader line “3 out of 4 units for Intelligence.” That meter is not an Intelligence Index integer. Speed is “Unknown out of 4 units.” Cost per Index task is not a dollar we can copy (the page leaves it unstated as a number). Chip omitted: aa stays null.
| Claim | Source | As of |
|---|---|---|
| OR catalog $3 / $15, web_search $0.005, 200k / 8k, created 1741312423, no tools param, hugging_face_id empty | OpenRouter catalog | 2026-08-19 |
| Official sonar-pro $3 / $15 per 1M plus $6 / $10 / $14 per 1,000 requests (low/medium/high) | Perplexity pricing.md | 2026-08-19 |
| AA page 200; printed $0.00 / $0.00; “3 out of 4 units for Intelligence” is a meter, not an Index integer; Speed Unknown; no copyable Index chip | Artificial Analysis Sonar Pro | 2026-08-19 |
| vals.ai perplexity_sonar-pro 404 | vals.ai Sonar Pro | 2026-08-19 |
| Official DeepSWE board: no exact Sonar Pro row | DeepSWE official board | 2026-08-19 |
| OpenRouter id perplexity/sonar-pro, context 200,000 | OpenRouter /api/v1/models | 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 Perplexity's model family with Sonar Reasoning Pro.
Because people search the name. The when-not block is the point of the card: this slug has no tools on the catalog we fetched.
Not as a hosted id. The 19 Aug 2026 public allowlist does not name sonar-pro. Use first-party or OpenRouter if you actually want search-grounded answers.
The page prints a 3-out-of-4 quality meter, not an Intelligence Index integer. Inventing 3 as the tile would be a fabricated score.
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 perplexity/sonar-pro.