fresh
Chinese frontier AI lab known for efficient open-weight reasoning and coding models. observed
Five evidence lanes (activity, capital, infra, hiring, ecosystem) as within-cohort percentiles, combined with evidence-breadth weights over the 60-company baseline cohort. The ± band widens with missing lanes — it measures what the score stands on. Scores publish at 2+ evidence lanes; this company's tape carries 1 (61 signals/60d), so no number is claimed.
Subdomains appearing in public certificate-transparency logs — infrastructure standing up, dated at detection. Intent labels are deterministic (token map), not guessed.
No standout signal or flag yet.
Cross-source signals that survived the scoring gauntlet or were hand-curated. ✓ verified = bound to an exact transcript passage with a voiceprint-confirmed speaker.
No single number decides. Layers can disagree — that's the point. New to this? →
The platform's own early signals for this company — hard-to-fake, and where one preceded a real raise, the measured lead.
This is a Chinese company actually called deep seek. And, um, they have this new model, uh, very small, only 7 billion parameters.
but they did move the market and hedge funds love to move the opposite direction because they're a hedge.
they did all, all the work to make an awesome AI model and then it happened to be so good that it changed the conversation.
keep in mind, deep seek is the subsidiary of a hedge fund. That hedge fund subsidiary just moved the market.
Maybe they released V3 back in December. R1 came out in January, but it's more like they just had some breakthroughs and pulled off some real innovation.
I think it's going to be open source all the way.
Great to get a kick in the ass. It's good for the industry. I think America still wins the day. I think open source still wins the day.
the rain on everyone's parade from deep seek is still raining.
it's called, uh, Janus pro seven B it makes images.
right before we came on air, uh, deep sea dropped a new model.
And then he, Aaron Levy points out that deep seek is cheap and also blowing up. And he says, all this points to the race is only going to continue to accelerate.
pro America technologists openly supporting a Chinese model that was trained off of leading us frontier models with chips to likely violate export controls. And according to their own terms of service, take us data back to China.
deep seas work a gift to the world
So what we have seen here is power of open source. The hosted version of deep seek R one has clear lines around what it can and cannot answer. Clearly president of China off list. Whereas the hosted version does not. What's cool though, is you don't have to use the web one. You can just download it. It's open source and run it yourself and so it strips away all that cruft.
software will become more efficient as we saw with r1 using maybe more commoditized hardware less generation of hardware better software better techniques etc
this idea that they spent six million dollars probably leaves out hundreds of thousands of illegally imported h100s via singapore that they're not putting into that cost
there's substantial evidence that what deep seek did here is they distilled the knowledge out of open ai's models and i don't think open ai is very happy about this and i think one of the things you're going to see over the next few months is our leading ag companies taking steps to try and prevent distillation and so we'll see if uh if the leading ag companies can prevent distillation by third-party companies that would definitely slow down some of these copycat models
Gray-market attention, not quality. 50+ HN mentions (peak 2091 points). Claim-shaped headlines are rumors until corroborated — a Tier-3 factual claim carries no evidentiary weight on its own. Risk drivers: heavy gray-market attention. ▲ 2091 DeepSeek v4 2026-04-24 ↗
| Attribute | Value | Kind | Source | Observed | State | Conf. |
|---|---|---|---|---|---|---|
common_name | DeepSeek | observed | notion | 2026-07-22 | normalized | 0.60 |
community.hn.max_points | 2091 | observed | artifact ↗ | 2026-08-16 | normalized | 0.20 |
community.hn.mentions | 50 | observed | artifact ↗ | 2026-08-16 | normalized | 0.20 |
community.hn.rumors | [] | observed | artifact ↗ | 2026-08-16 | normalized | 0.20 |
community.hn.top_story | {"url": "https://news.ycombinator.com/item?id=47884971", "title": "DeepSeek v4", "points": 2091, "created_at": "2026-04-24"} | observed | artifact ↗ | 2026-08-16 | superseded | 0.20 |
community.hn.top_story | {"url": "https://news.ycombinator.com/item?id=47884971", "title": "DeepSeek v4", "points": 2091, "created_at": "2026-04-24"} | observed | artifact ↗ | 2026-04-24 | superseded | 0.20 |
community.hn.top_story | {"url": "https://news.ycombinator.com/item?id=47884971", "title": "DeepSeek v4", "points": 2091, "created_at": "2026-04-24"} | observed | artifact ↗ | 2026-04-24 | superseded | 0.20 |
community.hn.top_story | {"url": "https://news.ycombinator.com/item?id=47884971", "title": "DeepSeek v4", "points": 2091, "created_at": "2026-04-24"} | observed | artifact ↗ | 2026-04-24 | normalized | 0.20 |
description | Chinese frontier AI lab known for efficient open-weight reasoning and coding models. | observed | notion | 2026-07-22 | normalized | 0.60 |
index.ai_tie | PURE_AI | observed | classify | 2026-07-25 | derived | 0.60 |
registry.signals | [{"claim": "Moonshot AI released Kimi K3, a 2.8-trillion-parameter open-source model claiming to be the world's largest, with benchmarks rivaling top proprietary systems from Anthropic and OpenAI.", " | observed | artifact ↗ | 2026-07-22 | superseded | 0.60 |
registry.signals | [{"claim": "Moonshot AI released Kimi K3, a 2.8-trillion-parameter open-source model claiming to be the world's largest, with benchmarks rivaling top proprietary systems from Anthropic and OpenAI.", " | observed | artifact ↗ | 2026-08-07 | normalized | 0.60 |
sector | emerging | derived | calc | 2026-08-08 | normalized | 0.50 |
signal.commentary | {"context": null, "run_ref": "01KYCR0FWHDNNY6R4HW6M5VH39", "signals": [{"quote": "there's substantial evidence that what deep seek did here is they distilled the knowledge out of open ai's models and | observed | artifact ↗ | 2025-01-29 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06N3PZEBJ0H0AS6V62YPGJ1", "signals": [{"quote": "this idea that they spent six million dollars probably leaves out hundreds of thousands of illegally imported h100s vi | observed | artifact ↗ | 2025-02-01 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06N3PZEBJ0H0AS6V62YPGJ1", "signals": [{"quote": "software will become more efficient as we saw with r1 using maybe more commoditized hardware less generation of hardwa | observed | artifact ↗ | 2025-02-01 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "So what we have seen here is power of open source. The hosted version of deep seek R one has clear lines around what i | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "deep seas work a gift to the world", "stance": "bullish", "signal_class": "endorsement", "speaker_name": "Mark Andrees | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "pro America technologists openly supporting a Chinese model that was trained off of leading us frontier models with ch | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "And then he, Aaron Levy points out that deep seek is cheap and also blowing up. And he says, all this points to the ra | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "right before we came on air, uh, deep sea dropped a new model.", "stance": "neutral", "signal_class": "fact_claim", "s | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "it's called, uh, Janus pro seven B it makes images.", "stance": "neutral", "signal_class": "fact_claim", "speaker_name | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "the rain on everyone's parade from deep seek is still raining.", "stance": "neutral", "signal_class": "assessment", "s | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "Great to get a kick in the ass. It's good for the industry. I think America still wins the day. I think open source st | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "I think it's going to be open source all the way.", "stance": "bullish", "signal_class": "prediction", "speaker_name": | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "Maybe they released V3 back in December. R1 came out in January, but it's more like they just had some breakthroughs a | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "keep in mind, deep seek is the subsidiary of a hedge fund. That hedge fund subsidiary just moved the market.", "stance | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "they did all, all the work to make an awesome AI model and then it happened to be so good that it changed the conversa | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M06NEEM9Q24N0RD8AXVEZM74", "signals": [{"quote": "but they did move the market and hedge funds love to move the opposite direction because they're a hedge.", "stance": | observed | artifact ↗ | 2025-01-27 | promoted | 0.50 |
signal.commentary | {"context": null, "run_ref": "01M0797B8660FSB38PKCG8Q8PH", "signals": [{"quote": "This is a Chinese company actually called deep seek. And, um, they have this new model, uh, very small, only 7 billion | observed | artifact ↗ | 2024-03-21 | promoted | 0.50 |
source.document.reference | {"kind": "video", "title": "Boom Supersonic's Milestones, OpenAI Ethics, and Startup Tools | E2078 — deepseek", "source_id": "01KXRSP95K9QAFKYJ3GAW4P1RP-m-6f7ac1823da81d2e", "extract_ref": "artifact:s | observed | artifact ↗ | 2026-07-25 | normalized | 0.95 |
source.document.reference | {"kind": "video", "title": "Is OpenAI Really Worth $300B? ElevenLabs’ Series C & YC’s Funding Focus | E2079 — deepseek", "source_id": "01KXRSP9560F36N17ZPWQQEJ5X-m-6f7ac1823da81d2e", "extract_ref": "a | observed | artifact ↗ | 2026-08-07 | normalized | 0.95 |
source.document.reference | {"kind": "video", "title": "DeepSeek Rattles Markets + SailPoint's Going Public… again? | E2077 — deepseek", "source_id": "01KXRSP962R51W9176S1TF3Z45-m-6f7ac1823da81d2e", "extract_ref": "artifact:sha2 | observed | artifact ↗ | 2026-08-07 | normalized | 0.95 |
source.document.reference | {"kind": "video", "title": "Twilio Bets Small, DeepSeek Scares & the GameOn Fraud | E2076 — deepseek", "source_id": "01KXRSP96G8W9HK6VY6A3BZK6Y-m-6f7ac1823da81d2e", "extract_ref": "artifact:sha256:633 | observed | artifact ↗ | 2026-08-07 | normalized | 0.95 |
source.document.reference | {"kind": "video", "title": "Microsoft consumes Inflection and AI Demos from Suno, Cohere, Deepseek VL and more! | E1918 — deepseek", "source_id": "01KXRSPB9AH8R8BCCWAAFTR24H-m-6f7ac1823da81d2e", "extr | observed | artifact ↗ | 2026-08-09 | normalized | 0.95 |
source.document.reference | {"kind": "video", "title": "AI Demos: Claude 3's Opus, Mistral, Groq Playground, EMO by Ali Baba | E1912 — deepseek", "source_id": "01KXRSPBBQTHW8C24ZM6HCDPTK-m-6f7ac1823da81d2e", "extract_ref": "arti | observed | artifact ↗ | 2026-08-09 | normalized | 0.95 |