Terza

Terza Inference is all in. The SN103 owner stake is 100% locked in a perpetual conviction lock. We're here to build Terza for the long run: one API for the best models, served by miners on Bittensor.

Verify on-chain: taostats.io/account/5HZJ7QuY…

ElevenLabs @ElevenLabs voices are now live on Terza Inference. All 7 speech models, from Eleven v4 to Flash, through the same API key you already use. From $0.0288 per 1K characters, 28% below the official price. Every voice in this video, all 18 of them, was generated through Terza. Try them in Studio terza.si

Meta's Muse Image is now on Terza Inference. Generate an image, then edit it with plain words. $0.0053 per image, 47% below Meta's price. Same API, same key. Try it in Studio terza.si

Terza Inference now accepts images on text models. Send screenshots and photos to Claude, GPT, Gemini, DeepSeek, GLM, Kimi and Qwen through the same API, in both OpenAI and Anthropic formats. Works in Claude Code too. Models that can't read images now return a clear error. No charge for rejected requests.

Context limits lifted on Terza Inference. Claude and GPT now take up to 1M input tokens and return up to 128K output tokens, with long responses streaming until they’re done. DeepSeek, Kimi, GLM, Qwen and Gemini follow over the next few hours. If you hit max_tokens or input too long errors last week, that's fixed. Thanks for the reports. 🙏

Long outputs are live on Terza Inference. max_tokens now goes up to 128K (131,072) on Claude and GPT models, and long responses stream for as long as they need. No more cut-offs mid-answer. Built for coding agents, long docs and big refactors.

DeepSeek starts billing Search & Fetch today ($0.75 / 1K calls), and its API still has peak-hour pricing. On Terza Inference, DeepSeek V4.1 Flash is one flat price, 24/7: $0.108 in / $0.432 out per 1M tokens That’s 28% below DeepSeek’s off-peak price and 64% below peak. Token-only pricing,no extra per-call fees. terza.si/pricing

Kling 4.0 is now live on Terza Inference 🎬 kling-video-v4: text-to-video and image-to-video, 3–30 s clips with native audio. 720p: $0.133/s 1080p: $0.1995/s About 15% below Kling’s official API price ($0.157 / $0.235). One Terza API key, billed per second, and you only pay for videos that are delivered. terza.si/pricing

New on Terza Inference: availability for every quote. Every route now shows: • Success, 24h: measured by our gateway on live requests • Uptime, 24h: reported by the miner serving it Choose by price and reliability. Failover covers the rest. terza.si/pricing

New on Terza Inference ⚡ 🖼️ grok-imagine-image-2.0: $0.0318/image, up to 60% below xAI 🧠 kimi-k3.1-pro: $0.72/$3.60 per 1M tokens, 64% below Moonshot 🎧 qwen3.8-omni-flash: text, image & audio in, 40% below Alibaba One API key. Pay per request. terza.si

More ways to top up. We've partnered with AllScale to bring fiat top-ups to Terza Inference. @allscaleio Prefer TAO? You can now top up with TAO, too. One balance. More ways in. terza.si

One model. Multiple routes. Miners provide inference. Terza brings access together through one API key. Choose the model your product needs, without wiring up a separate integration for every route. terza.si

Top up Terza Inference this week and get 20% extra credit. Any amount, every top-up — added to your balance automatically when the payment lands. Ends Nov 4, 10:59 AM ET (7:59 AM PT). terza.si

More room to build. Same models, lower bills. Terza pricing per 1M tokens, input and output: vs SayGM • GPT-6.1 Sol: 94% less • GPT-6 Astra: 94% less • Claude Sonnet 5.5: 86% less • Gemini 3.8 Flash: 85% less • Claude Opus 5.5: 73% less • Kimi K3: 46% less vs Chutes • DeepSeek V4 Flash 0731: 64% less • Kimi K3: 41% less • GLM-5.2: 32% less vs engy • GLM-5.3 Flash: 60% less • Kimi K3: 44% less • GLM-5.3: 14% less vs Official Pricing • GPT-6.1 Sol: 96% less • GPT-6 Astra: 96% less • Claude Sonnet 5.5: 89% less • Gemini 3.8 Flash: 91% less • Claude Opus 5.5: 89% less • Kimi K3: 64% less • GLM-5.3 Flash: 64% less • GLM-5.2: 40% less • GLM-5.3: 40% less Check model pricing before calling. terza.si

Jev by TypeSafe is now live in the Terza Playground. @typesafeai Classify. Route. Score. Structured decisions, with input pricing 52% below TypeSafe's official rate. Try Jev with your Terza API key. terza.si

Terza is live on mainnet. Many models. One API. Access miner-provided inference through a single Terza key. Built on Bittensor, SN103. Watch the launch film. Then build something. terza.si

Meet Terza. Model access, through one API. Built on Bittensor. Our website is live and the network is running on mainnet terza.si