SimplyTao Enters Open Beta | Weekly Bittensor Update
This week brought a major milestone for SimplyTao as the platform officially opened its doors to everyone with no whitelist required. Meanwhile, the Opentensor Foundation announced an on-site Bittensor hackathon in San Francisco with up to 1,000 TAO in funding for winning teams. On the subnet side, Ridges (SN62) launched Ridgeline in open beta, Chutes (SN64) spotlighted a reasoning model that undercuts GPT-5 pricing by over 90%, Score (SN44) expanded beyond football with its first new vision task, and IOTA (SN9) rolled out multi-run support for decentralized training. This week’s top gainer is SOMA (SN114), up over 157%. $TAO closed the week with a +4.97% gain, trading between $173 and $198.
Big Events in the Bittensor Ecosystem
SimplyTao Is Now in Open Beta
Our platform has officially entered Open Beta! SimplyTao is now accessible to everyone without a whitelist or invite code. If you’re interested in the Bittensor ecosystem, you can create an account and start exploring subnet tokens right away.
We’re building SimplyTao together with the community, so your early feedback plays a real role in shaping what comes next. More subnets, additional payment gateways, staking options and new features are already in the pipeline. A full roadmap is dropping next week.
To celebrate the launch, we’re offering a 30% cashback for everyone who signs up with the referral code SIMPLYT01.
Bittensor On-Site Hackathon in San Francisco
The Opentensor Foundation has announced an on-site Bittensor hackathon in San Francisco. The event takes place at Frontier Tower, 995 Market St, running from March 11 (9:00 AM) to March 15 (1:00 PM). Teams will compete to build a sovereign AI layer as a Bittensor subnet.
Applications are now open. Judging criteria focus on market demand, miner task design, validator scoring, incentive mechanics, economic viability and technical feasibility. In short, judges want to see subnets that solve real problems and can sustain themselves on mainnet.
The top 3 teams will compete for a strong prize pool. Rewards include up to 1,000 TAO in discretionary funding, an investor pitch with Unsupervised Capital, 50 TAO toward mainnet registration from Crucible Labs, $5,000 in GPU compute credits from Basilic AI and a cash bounty split of $1,000 / $600 / $400. Winners also receive post-event launch support.
Judges include representatives from BitMind, Distributed State and other ecosystem builders. Partners behind the event are Funding Commons, Protocol Labs, Frontier Tower and Redwood North TAO.
General Tensor Launches New Website
General Tensor has officially launched GeneralTensor.io. The new website brings together everything the team is building across the Bittensor ecosystem in one place. General Tensor describes itself as a vertically integrated producer of digital intelligence, operating across blockchain, AI and data centers.
The site showcases their full portfolio stack, including applications like Project Rubicon and 0xMarkets, their validator operations (formerly known as RoundTable21), a subnet incubator with projects like Cartha (SN35), TPN (SN65), Red Team (SN61), Trishool (SN23) and Synth (SN50), as well as their mining and data center divisions. The validator currently holds over 400,000 TAO in delegation from more than 6,000 delegators, ranking as the 6th largest by delegation on the network.
Subnet Updates
Ridgeline by Ridges (SN62) Is Now in Open Beta
Ridges (SN62) has launched Ridgeline, a product layer that brings the subnet’s AI coding agents directly to developers. The tool integrates with GitHub, allowing developers to assign issues to an agent that reads the repository, generates patches and tests them automatically. Ridgeline runs inference through Targon (SN4) and Chutes (SN64), making it a strong example of cross-subnet composability on Bittensor. The open beta has no paywall, and every new account receives 10 free credits.
Read our full breakdown of Ridgeline by Ridges (SN62) here.
Chutes (SN64) Spotlights Qwen3-235B as a Budget Reasoning Powerhouse
Chutes (SN64) is highlighting Qwen3-235B-A22B Thinking-2507 as the best price-to-performance reasoning model currently available on its platform. The model offers 262K context, supports tools, structured outputs and JSON mode, and delivers reasoning quality competitive with GPT-5 on math, logic and multi-step tasks.
The pricing difference is significant. GPT-5 runs at $1.25/M input and $10.00/M output, while Qwen3-235B Thinking on Chutes costs just $0.11/M input and $0.60/M output. That translates to roughly 91% cheaper input and 94% cheaper output. For a team processing 10M tokens per day, the monthly savings come out to around $3,500 compared to OpenAI.
The model fits best in workflows that require complex reasoning chains, code review at scale, scientific computation and research synthesis from long documents. It runs through an OpenAI-compatible API, so it can drop into any existing workflow.
Score (SN44) Introduces Miner Starter Pack and Expands Beyond Football
Score (SN44) has released its first Miner Starter Pack, marking the arrival of the first non-football vision task on the subnet. This is a significant shift, as SN44 begins expanding into a broader set of real-world vision elements beyond its original sports focus.
The release also introduces a new standardized operating model for how future tasks roll out on the subnet. Each new task gets announced alongside a Miner Starter Pack so miners can prepare. A manifest is then automatically generated to define the evaluation framework.
From there, a test phase begins with 1% emissions, which can scale up to 10% depending on task complexity. Miners also get console access to monitor and interact with tasks in real time. Test Phase #1 is set to begin in mid-March.
IOTA (SN9) Adds Multi-Run Support to Training at Home
IOTA (SN9) has rolled out multi-run support for its Training at Home program. Three new concurrent open public runs are now active, with a total pool size of 450 nodes. The team reports that their newest systems are training faster, producing better losses earlier in the process, and showing more resilience to compute node churn.
The next major upgrade is coming soon. Peer-to-peer communication is set to release next week, which should unlock faster data exchange between nodes, support for deeper models and accelerate research output overall. Progress can be tracked on the IOTA dashboard.
TAO Market Update
Price: $173.06 – $198.39
Weekly: +4.97%
Ranking: #39
Market Cap: $1.87B
24h Volume: $137M
Top Gainer Subnet: SOMA (SN114)
SOMA (SN114) is a Bittensor subnet that brings MCP (Model Context Protocol) servers into the decentralized ecosystem. It enables AI models to securely connect with external tools, databases and APIs through a standardized interface. Miners compete in weekly cycles to deliver the best-performing algorithms for specific tasks.
The first competition focuses on context compression, which reduces inference costs and extends memory windows for AI agents. Submissions open on March 12 at 1pm UTC and run through March 16, followed by an evaluation phase until March 19 when the winner is selected. The subnet is built by Dendrite HQ and aims to become the go-to infrastructure layer for decentralized AI tooling across the Bittensor network.
Weekly change: +157.55%
Price: $3.468732
Market Cap: $846.73K
Volume (24h): $379.21K
