BullTensor
TAO $277.90
Market cap · 30d
$42.2M τ152K
13.78% / 24h ↗
24h volume
$0 —
Daily, 24 days
Supply
25.6%
5.37M ד
in circulation, out of 21M max
  • In circulation 5.37M ד · 25.6%
  • In the pool 936K ד · 4.5%
  • Not emitted yet 14.7M ד · 70.0%
Locks
$2.83M 6.7% of circulating supply
  • Owner lock 5,000.00 ד · 0.1%
  • Other holders (2) 356K ד · 6.6%
  • Not locked 5.01M ד · 93.3%
∞ Perpetual lock 15.2K ד · ↘ Decaying lock 346K ד

What is SayGM?

In short
  • SayGM is one key to use Claude, GPT, Gemini and open AIs, instead of an account with each provider.
  • Frontier models cost less there than from their official maker (up to 45% less right now).
  • Participants of the Bittensor network compete to offer the best price, on subnet 28 (SN28).

To use Claude, GPT and Gemini in an app, you normally need an account and a bill with Anthropic, OpenAI and Google. SayGM puts everything behind a single key: you pick the model, SayGM handles the rest, and you often pay less than the official price.

It also offers open AIs (DeepSeek, GLM, Qwen, Kimi…) and "confidential" versions. SayGM is part of decentralized AI inference and receives 3.26% of the network's rewards.

How it works

  1. 01
    One key for every model

    You send your questions to SayGM and name the model you want: Claude, GPT, Gemini or an open model.

  2. 02
    Participants make offers

    The subnet's miners offer to handle requests and compete on price; SayGM picks the cheapest offer that meets its criteria.

  3. 03
    You pay less

    Competition pushes the price below the maker's official rate: for example 13.2% less on Claude Sonnet 5.5.

Going further: the technical details

Displayed prices use, in order: recent finalized-epoch results, then current miner offers, then SayGM's own list price. The "official price" is the maker's published rate, returned by SayGM's API.

Models ending in "TEE" run in a Trusted Execution Environment, confidential-computing hardware that isolates the model and your prompts from the machine's operator.

Docs and API keys on saygm.com; model and price list:

GET https://api.saygm.com/v1/models

Owner on-chain revenue

What the subnet's team earns: 18% of the tokens the subnet creates every day go to it. It's its on-chain salary, visible to everyone.

Per day (current)
$10,175.65
τ36.62 · 1,295.97 ד
Last 30 days
$241K
τ850.47
Last 90 days
$522K
τ2,159.69

Daily revenue in USD (owner emissions × alpha price × TAO/USD of the day).

Alpha locks

on-chain state 42 minutes ago

When a team locks its own tokens, it can't sell them: it's a sign of confidence in its project. Other holders can lock too, behind the team's key.

Locked by the owner
5,000.00 ד
0.1% of circulating supply · $39.3K
Type
∞ Perpetual
Emission auto-lock disabled
Total on the owner hotkey
361K ד
∞ 15.2K · ↘ 346K
Wallets locking
3
on the owner hotkey
ColdkeyHotkeyAlpha locked % circulatingTypeLast update
5FqACM…QzsThe owner hotkey 345,850.80 ד 6.44% ↘ Decaying #9218027
5HMtFE…eS8CCx owner hotkey 5,192.81 ד 0.10% ∞ Perpetual #8865642
5CA6rw…RQxm45 Owner owner hotkey 5,000.00 ד 0.09% ∞ Perpetual #9244023
How it's measured

Alpha locked via Lock / OwnerLock. A perpetual lock (∞) never unlocks; a decaying lock (↘) is released progressively (~130 days). An owner locking its own alpha signals confidence in the project.

Buybacks

Some teams use their revenue to buy back their own token and destroy it ("buyback & burn"): fewer tokens in circulation, like a company buying back its shares.

Alpha burned 30d
—
Buyback & burn
…
Manual burns
…
Emission burns
…
Burned by the owner (30d)
0.00 ד

The split between buyback & burn, manual burns and emission burns appears once the events of the period are indexed (history scan in progress).

60 days excluded: missing reading, or the counter moved more than the subnet can burn (e.g. runtime migration rewriting it).

TAO spent 30d
τ0
$0
TAO spent (all indexed)
τ0
Alpha burned 30d
0.00 ד
Operations
0

No buyback indexed yet.

How it's measured

On-chain events: AddStakeBurn (buyback & burn), alpha purchases (StakeAdded) and burns/recycles made by this subnet's declared treasury wallets.

Claude, GPT and Gemini for less (live)

● SayGM API responding · 585 ms · checked 19 minutes ago · 24h uptime 100.0%

Each frontier model's price from its maker, next to its price on SayGM. Prices in $ per million tokens, input / output (a million tokens is about 750,000 words).

Frontier models offered
34
Biggest discount
▼ 45.0%
Makers
Anthropic, Google, OpenAI
ModelMakerOfficial priceSayGMSaving
Claude Fable 5 Anthropic $10 / $50 $8.24 / $41.2 ▼ 17.6%
Claude Fable 5.1 Anthropic $10 / $50 $8.24 / $41.2 ▼ 17.6%
Claude Opus 4.7 Anthropic $5 / $25 $4.12 / $20.6 ▼ 17.6%
Claude Opus 4.8 Anthropic $5 / $25 $4.12 / $20.6 ▼ 17.6%
Claude Opus 5 Anthropic $5 / $25 $4.12 / $20.6 ▼ 17.6%
Claude Opus 5.5 Anthropic $4 / $20 $3.472 / $17.36 ▼ 13.2%
Claude Sonnet 4.6 Anthropic $3 / $15 $2.472 / $12.36 ▼ 17.6%
Claude Sonnet 5 Anthropic $2 / $10 $1.648 / $8.24 ▼ 17.6%
Claude Sonnet 5.5 Anthropic $2 / $10 $1.736 / $8.68 ▼ 13.2%
Claude Haiku 4.5 Anthropic $1 / $5 $0.824 / $4.12 ▼ 17.6%
Claude Haiku 5.5 Anthropic $0.1 / $0.5 $0.085 / $0.425 ▼ 15.0%
Gemini 3.1 Pro Preview Google $2 / $12 $1.3 / $7.8 ▼ 35.0%
Gemini 3.5 Flash Google $1.5 / $9 $0.975 / $5.85 ▼ 35.0%
Gemini 3.6 Flash Google $0.75 / $3.75 $0.4875 / $2.4375 ▼ 35.0%
Gemini 3.7 Flash Google $0.75 / $3.75 $0.4875 / $2.4375 ▼ 35.0%
Gemini 3.8 Flash Google $0.75 / $3.75 $0.4125 / $2.0625 ▼ 45.0%
Gemini 3.1 Flash Image Google $0.5 / $3 $0.325 / $1.95 ▼ 35.0%
Gemini 3.5 Flash Lite Google $0.3 / $2.5 $0.195 / $1.625 ▼ 35.0%
Gemini 3.1 Flash Lite Google $0.25 / $1.5 $0.1625 / $0.975 ▼ 35.0%
Gemini 3.1 Flash Lite Image Google $0.25 / $1.5 $0.1625 / $0.975 ▼ 35.0%
GPT-5.5 Pro OpenAI $30 / $180 $22.35 / $134.1 ▼ 25.5%
GPT-6 Astra OpenAI $10 / $50 $7.45 / $37.25 ▼ 25.5%
GPT-5.5 OpenAI $5 / $30 $3.725 / $22.35 ▼ 25.5%
GPT-5.6 Sol OpenAI $4 / $20 $2.98 / $14.9 ▼ 25.5%
GPT-5.4 OpenAI $2.5 / $15 $1.8625 / $11.175 ▼ 25.5%
GPT-5.6 Terra OpenAI $2 / $12 $1.49 / $8.94 ▼ 25.5%
o3 OpenAI $2 / $8 $1.49 / $5.96 ▼ 25.5%
GPT-6 Sol OpenAI $2 / $10 $1.49 / $7.45 ▼ 25.5%
GPT-6.1 Sol OpenAI $2 / $10 $1.49 / $7.45 ▼ 25.5%
o4-mini OpenAI $1.1 / $4.4 $0.8195 / $3.278 ▼ 25.5%
GPT-5.4 mini OpenAI $0.75 / $4.5 $0.5588 / $3.3525 ▼ 25.5%
GPT-5.4 nano OpenAI $0.2 / $1.25 $0.149 / $0.9313 ▼ 25.5%
GPT-5.6 Luna OpenAI $0.2 / $1.2 $0.149 / $0.894 ▼ 25.5%
GPT-6 Luna OpenAI $0.1 / $0.5 $0.0745 / $0.3725 ▼ 25.5%

Saving computed on a 75% input / 25% output mix. Off-chain data from SayGM's public API, refreshed every 10 minutes.

Open and confidential AIs

Open-source models (DeepSeek, GLM, Qwen, Kimi…), some running on confidential hardware (TEE).

ModelMakerInput $/1M tokensOutput $/1M tokens
Mistral Nemo TEE Mistral $0.0233 $0.0929
Nemotron 3 Nano Omni 30B TEE Nvidia $0.0233 $0.0929
Qwen3.8-27B Qwen $0.045 $0.27
gpt-oss-20b Openai $0.0665 $0.285
Qwen3.6-35B-A3B Qwen $0.0771 $0.4618
DeepSeek-V4-Flash-0731 Deepseek $0.0866 $0.2599
Qwen3-32B TEE Qwen $0.0988 $0.3952
Gemma 4 31B turbo TEE Google $0.114 $0.3515
GLM-5.3-Flash Zai $0.132 $0.44
GLM-5.3-Flash TEE Zai $0.135 $0.45
DeepSeek-V4.1-Flash Deepseek $0.14 $0.5599
DeepSeek-V4.1-Flash TEE Deepseek $0.27 $1.08
Qwen3-235B-A22B-Thinking-2507 TEE Qwen $0.284 $1.1359
Qwen3.6-27B TEE Qwen $0.285 $1.9
DeepSeek-V4-Flash-0731 TEE Deepseek $0.418 $1.254
Qwen3.5-397B-A17B TEE Qwen $0.4275 $2.85
Kimi-K2.6 TEE Moonshot $0.475 $2.7075
GLM-5.2 Zai $0.5163 $1.6227
Ornith 1.5 397B Ornith $0.7 $2.2
Ornith 1.5 397B TEE Ornith $0.7 $2.2
GLM-5.3 Zai $0.784 $2.464
DeepSeek-V3.2 TEE Deepseek $0.95 $0.95
GLM-5.2 TEE Zai $1.26 $3.96
GLM-5.3 TEE Zai $1.26 $3.96
GLM-5.1 TEE Zai $1.33 $4.18
Kimi K3 Moonshot $1.5 $7.5
Kimi-K3 TEE Moonshot $2.85 $14.25
MiMo-V2.6-Pro-UltraSpeed Xiaomi $3.915 $7.83
MiMo-V2.6-Pro-UltraSpeed TEE Xiaomi $3.915 $7.83

Source: api.saygm.com/v1/models, refreshed every 10 minutes. Off-chain data, unlike the rest of this page.

SayGM vs the centralized market

SayGM is cheaper than the median centralized provider on 46 of the 50 models also sold on the traditional market.

Model SayGM Market median vs median Official API Cheapest centralized
qwen3.8-27b $0.1013 $0.7375 (19) ▼ 86% cheaper $0.9563 Alibaba $0.3675 Tenstorrent, fp4
glm-5.2 $0.7929 $1.9663 (25) ▼ 60% cheaper $2.15 Z.AI $0.258 Baidu, fp8
qwen3-235b-a22b-thinking-2507-tee $0.4969 $1.2125 (1) ▼ 59% cheaper — $1.2125 Venice, fp8
qwen3.6-35b-a3b $0.1733 $0.3438 (10) ▼ 50% cheaper — $0.2125 Darkbloom, fp4
kimi-k3 $3 $5.8 (19) ▼ 48% cheaper $6 Moonshot AI $3.0475 InferenceNet, fp4
gemini-3.8-flash $0.825 $1.5 (2) ▼ 45% cheaper $0.75 Google AI Studio $0.75 Google AI Studio
deepseek-v4-flash-0731 $0.13 $0.231 (24) ▼ 44% cheaper — $0.09 DeepInfra, fp8
deepseek-v4.1-flash $0.245 $0.4213 (29) ▼ 42% cheaper $0.2625 DeepSeek $0.1125 Decart, fp4
glm-5.3 $1.204 $2.0188 (33) ▼ 40% cheaper $2.15 Z.AI $0.3053 Baidu, fp8
gemini-3.1-flash-lite $0.3656 $0.5906 (2) ▼ 38% cheaper $0.2813 Google $0.2813 Google
gemini-3.5-flash-lite $0.5525 $0.8925 (2) ▼ 38% cheaper $0.425 Google $0.425 Google
gemini-3.1-flash-image $0.7313 $1.125 (2) ▼ 35% cheaper $1.125 Google AI Studio $1.125 Google AI Studio
gemini-3.1-flash-lite-image $0.3656 $0.5625 (2) ▼ 35% cheaper $0.5625 Google $0.5625 Google
gemini-3.1-pro-preview $2.925 $4.5 (2) ▼ 35% cheaper $2.25 Google $2.25 Google
gemini-3.5-flash $2.1938 $3.375 (2) ▼ 35% cheaper $1.6875 Google $1.6875 Google
gemini-3.6-flash $0.975 $1.5 (2) ▼ 35% cheaper $0.75 Google $0.75 Google
gemini-3.7-flash $0.975 $1.5 (2) ▼ 35% cheaper $0.75 Google $0.75 Google
qwen3.6-27b-tee $0.6888 $1.04 (6) ▼ 34% cheaper $1.0125 Alibaba $0.8425 Reka
gpt-5.4 $4.1906 $6.1875 (3) ▼ 32% cheaper — $2.8125 OpenAI
gpt-5.5 $8.3813 $12.375 (3) ▼ 32% cheaper — $5.625 OpenAI
gpt-5.6-luna $0.3353 $0.495 (3) ▼ 32% cheaper — $0.225 OpenAI
gpt-5.6-terra $3.3525 $4.95 (3) ▼ 32% cheaper — $2.25 OpenAI
gpt-6-astra $14.9 $22 (3) ▼ 32% cheaper — $10 OpenAI
gpt-6-luna $0.149 $0.22 (3) ▼ 32% cheaper — $0.1 OpenAI
gpt-6-sol $2.98 $4.4 (3) ▼ 32% cheaper — $2 OpenAI
gpt-6.1-sol $2.98 $4.4 (3) ▼ 32% cheaper — $2 OpenAI
kimi-k2.6-tee $1.0331 $1.4375 (16) ▼ 28% cheaper $1.7125 Moonshot AI $0.7792 Baidu, fp4
gpt-5.4-mini $1.2572 $1.6875 (2) ▼ 26% cheaper — $0.8438 OpenAI
gpt-5.4-nano $0.3446 $0.4625 (2) ▼ 26% cheaper — $0.2313 OpenAI
gpt-5.6-sol $5.96 $8 (3) ▼ 26% cheaper — $2 OpenAI
o3 $2.6075 $3.5 (1) ▼ 26% cheaper — $3.5 OpenAI
o4-mini $1.4341 $1.925 (1) ▼ 26% cheaper — $1.925 OpenAI
claude-opus-5 $8.24 $11 (5) ▼ 25% cheaper — $10 Claude Platform on AWS
claude-sonnet-5 $3.296 $4.2 (5) ▼ 22% cheaper — $4 Claude Platform on AWS
qwen3.5-397b-a17b-tee $1.0331 $1.3 (10) ▼ 21% cheaper $0.8775 Alibaba $0.8775 Alibaba
claude-fable-5 $16.48 $20 (5) ▼ 18% cheaper — $20 Claude Platform on AWS
qwen3-32b-tee $0.1729 $0.205 (3) ▼ 16% cheaper — $0.13 DeepInfra, fp8
mimo-v2.6-pro-ultraspeed $4.8938 $5.4375 (1) ▼ 10% cheaper — $5.4375 Xiaomi, fp8
mimo-v2.6-pro-ultraspeed-tee $4.8938 $5.4375 (1) ▼ 10% cheaper — $5.4375 Xiaomi, fp8
glm-5.3-flash $0.209 $0.2234 (28) ▼ 6% cheaper $0.2375 Z.AI $0.065 InferenceNet, fp4
glm-5.3-flash-tee $0.2138 $0.2234 (28) ▼ 4% cheaper $0.2375 Z.AI $0.065 InferenceNet, fp4
glm-5.3-tee $1.935 $2.0188 (33) ▼ 4% cheaper $2.15 Z.AI $0.3053 Baidu, fp8
glm-5.1-tee $2.0425 $2.0913 (12) ▼ 2% cheaper $2.15 Z.AI $1.4814 Baidu, fp8
kimi-k3-tee $5.7 $5.8 (19) ▼ 2% cheaper $6 Moonshot AI $3.0475 InferenceNet, fp4
glm-5.2-tee $1.935 $1.9663 (25) ▼ 2% cheaper $2.15 Z.AI $0.258 Baidu, fp8
gpt-5.5-pro $50.2875 $50.625 (1) ▼ 1% cheaper — $33.75 OpenAI
deepseek-v4.1-flash-tee $0.4725 $0.4213 (29) ▲ 12% pricier $0.2625 DeepSeek $0.1125 Decart, fp4
gpt-oss-20b $0.1211 $0.06 (10) ▲ 102% pricier — $0.036 Darkbloom, fp8
deepseek-v3.2-tee $0.95 $0.4375 (12) ▲ 117% pricier — $0.234 GMICloud, fp8
deepseek-v4-flash-0731-tee $0.627 $0.231 (24) ▲ 171% pricier — $0.09 DeepInfra, fp8

Blended price in USD per million tokens (3 input tokens for 1 output token, a typical chat/agent mix). Market = centralized providers selling the same model on OpenRouter (number of providers in brackets), refreshed hourly; Bittensor-based providers are excluded. Very cheap offers often use stronger quantization (fp4): check quality before switching.

SayGM FAQ

What is SayGM?
SayGM is Bittensor subnet 28: a single API to use Claude, GPT, Gemini and open-source models. Bittensor miners compete on the price of requests, so you can pay less than the makers' official rates.
Is SayGM cheaper than Anthropic, OpenAI or Google?
On frontier models, yes right now: up to 45.0% below the maker's official price (13.2% on Claude Sonnet 5.5). The table above compares every model live.
Which models does SayGM offer?
63 models right now, including 34 frontier models (Anthropic's Claude, OpenAI's GPT, Google's Gemini) and open-source models such as DeepSeek, GLM, Qwen or Kimi, some in a confidential (TEE) version.
Is SayGM down?
SayGM's API answered our last check, 19 minutes ago (585 ms).
What is SayGM's token?
The alpha token of Bittensor subnet 28 (symbol ד), traded against TAO in the subnet's on-chain pool: currently τ0.0282540 ($7.85), for a market cap of $42.2M.