Subnets / SN67
Harnyx SN67
Harnyx (subnet 67) is a Bittensor subnet held by 1,239 unique coldkeys. Its liquidity pool holds 2.4k τ, with the alpha token trading at 0.0027 τ. Roughly 13% of tracked holders are in profit — most bought in above today's price. Seller exhaustion is elevated: 96% of recent sellers have emptied their wallets, thinning the supply overhang. Full insider flow, whale tracking and the complete proprietary factor set for Harnyx are available to SubnetStats subscribers.
ALPHA PRICE
0.0027 τ
IN TAO
HOLDERS
1,239
UNIQUE COLDKEYS
POOL DEPTH
2.4k τ
TAO IN POOL
7D MOMENTUM
-4.0%
ALPHA vs TAO
Project activity
SN67 now crowns two champions.
The main title still goes to the best deep research: accurate answers in a full report, every claim traced to its source. The new title rewards accuracy on its own. 🧵
$TAO #SN67 https://t.co/e1LSTkTVSB
Anthropic gave analysts an AI agent secretly told to push one side of each decision. They sided with it 86% of the time. None suspected the push was planted.
Harnyx runs deep research on an open network: anyone can compete, and no single vendor owns the preference.
$TAO #SN67
SN67 miners can now call @OpenRouter rerank models as a tool.
Search and fetch still find the pages. Rerank sorts the candidates before you burn an llm_chat call on the wrong ones.
Paid models only. Same OpenRouter key.
$TAO #SN67 https://t.co/sAZucIylcZ
SN67 miners can now call decision models from @reaxlabs (SN92).
A research run is full of small decisions — which source to trust, which path to follow, when to stop. Those snap calls shouldn't burn a frontier model.
Research on one subnet. Decisions on another.
$TAO #SN67 htt
SN67 miners can now call @Cloudflare Workers AI natively.
Chat models through llm_chat. Clef and Clairic through decision_query — snap calls and extraction on the same tools they already use.
Published rates only. Promo credits don’t shrink the task budget.
$TAO #SN67 https://
Major update: we've removed our model allowlist on @OpenRouter.
SN67 miners can now build with any paid LLM or decision model there. Frontier labs and open models, one competition.
We no longer pick the models. The swarm does. 🧵
$TAO #SN67 https://t.co/wjpW70eldA
When you hand research to AI, the answer comes back clean and confident. You often can’t tell which numbers came from a source, which were guessed, or what it never found.
Five questions to check any research output before you act on it: https://t.co/3AUgguZSiZ
Rolling out on SN67 from today: a task's time limit depends on how hard it is.
Two research agents try each task before miners do.
- Neither solves it: 300s
- One does: 200s
- Both do: 100s
Fast mode: 70% of each.
Real API use means easy questions come back fast.
$TAO #SN67 ht
Decision models make snap calls. Span-02 from @RespanAI can also stop and think, and still returns its best call when cut off.
Hundreds of SN67 agents now learn which calls deserve a second thought.
Also new: Drex 2.0 (@NaceAI), Step 5 Preview (@StepFun_ai).
$TAO #SN67
Muse Spark Open from @AIatMeta is now in the SN67 miner toolchain.
It's built for long-horizon, multi-step work with a 1M-token context. That's the shape of a deep research run: many sources, many steps, one answer that has to hold together.
Now the swarm tests it.
$TAO #SN67
Yesterday @deepseek_ai shipped a new Flash model and opened its in-house search engine to everyone.
Both are now tools every SN67 miner can put in their research agent. 🧵
$TAO #SN67
Expand Fetch from @getexpandai is now a page-fetching tool for every SN67 miner.
Each fetch returns the page as it looks, tables intact, plus the hidden data behind it as JSON. Miners can cite each separately, so every claim shows exactly what the agent saw.
$TAO #SN67
Last time, a research task shipped if two frontier models both missed it. That bar is now much stricter.
We also upgraded the agent that writes the questions, so they're harder before they ever reach that test.
More room at the top to tell the best agents apart.
$TAO #SN67
SN67 miners can now run their agents on @say_gm_, the LLM gateway built on Bittensor SN28.
Every model is up for auction. Operators bid to serve it, and prices land below list.
Our miners compete on research. Theirs compete on inference.
$TAO #SN67 https://t.co/rhsQdpGQbU
A research run is full of small decisions: which source to trust, which path to follow, when to stop. Paying a frontier model for each is the waste. Decision models are now a tool the SN67 swarm can call for that step. Cheaper per decision, cheaper deep research.
$TAO #SN67
The hard part of deep research used to be answering the question. On SN67, it's becoming writing one the swarm can't already solve.
We keep raising the bar; miners keep clearing it. This might be one of the last raises before frontier. 🧵
$TAO #SN67 https://t.co/z78zWcqDSB
MiMo-V2.6 is now in the SN67 miner toolchain. Open weights, a 1M-token context, and a Pro variant that serves up to 20x faster. Every miner in the swarm can now put it to work on deep research.
$TAO #SN67
Harnyx keeps evolving.
7.5% → 70% in under six months.
Weekly median fully-correct rate on DeepSearchQA (20-question samples—not the full benchmark).
Behind that progress: 183 batches and 31,648 unique miner submissions since April.
Still evolving.
More benchmarks and run-level
The main round is where a challenger earns the title on SN67, so we widened it. Every artifact that beats the champion now advances into the full contest, and a bigger field is what pushes the frontier. 🧵
$TAO #SN67 https://t.co/TJkkWckAjG
DeepSeek-V4.1-Flash is now in the SN67 miner toolchain. Every miner in the swarm can call it in their agents.
New architecture, native multimodal, and fast. One more route for the swarm to reach a good answer.
$TAO #SN67
Fast mode now covers speed-only requests, so the default path can be held to a stricter bar.
The evaluator checks every citation index deterministically. A response with any invalid citation fails, so nothing with a broken citation reaches the caller.
$TAO #SN67
Novelty checking on SN67 now has to clear two independent comparisons instead of one. That closes a gap where a submission could hide an older mechanism under enough new structure to still read as original. 🧵
$TAO #SN67 https://t.co/5mPv5UHTnJ
hy4-preview is now in the SN67 miner toolchain. Every miner in the swarm can put it to work on deep research.
$TAO #SN67
Introducing fast mode on SN67.
Two ways to run deep research now: fast and cheap when you just need the answer, fully traceable when every claim has to be verifiable. Same miner swarm, your call on the tradeoff. 🧵
$TAO #SN67 https://t.co/SvMntxHerI
GLM-5.3-Flash is now in the SN67 miner toolchain.
Open weights, natively multimodal, a 1M-token context, at a fraction of the last generation's cost. Every miner in the swarm can now put it to work on deep research.
$TAO #SN67
SELLERS DRAINED
96%
28D SELLERS EMPTIED OUT
CAPITULATION
+0.82
BOUNCE COMPOSITE (z)
HEAT 7D
0.11×
VOLUME ÷ POOL DEPTH
HOLDERS IN PROFIT
13%
TRACKED WALLETS · ENTRY BELOW PRICE
WHAT'S CARRYING THE PRICE · 30D—emission rank #67 · below the bar (rank 32)
30D RETURN
-20.2%
ALPHA vs TAO
WITHOUT CHAIN BUYS
-20.2%
SAME 30 DAYS, PROTOCOL BID REMOVED
CHAIN SHARE OF BUYING
—
CHAIN BUYS vs TRADERS' NET
CHAIN BUYS / DAY
0 τ
PROTOCOL BUYING ITS ALPHA
See who's buying and selling Harnyx — every trade classified miner / validator / owner / outsider, plus holder cost basis and the full 23-factor screener.
OPEN THE INSIDER TAPE →