MagmaRouter vs Together AI
Together AI is a full AI cloud with 200+ open models, fine-tuning, dedicated endpoints and GPU clusters, all on its own infrastructure. MagmaRouter is a much smaller gateway that offers only a short line of uncensored Xploded models, with hard spend caps on each key.
Facts checked October 2026 against Together AI’s own site, docs and policies.
| MagmaRouter | Together AI | |
|---|---|---|
| What it is | Hosted gateway for one uncensored, tool capable model line (Xploded). | An AI cloud offering serverless inference on open models, batch inference, provisioned throughput, dedicated endpoints, fine-tuning and GPU clusters. |
| Pricing | Pay as you go, $5 minimum top-up, no subscription required. | Serverless is per token, input and output priced separately, for example GLM-5.3-Flash at $0.15 in and $0.50 out, Qwen3.8 Flash at $0.09 in and $0.28 out per 1M tokens. Fully prepaid with a $5 minimum credit purchase; their billing docs say free trials are not offered. Dedicated endpoints are billed per GPU hour, for example H100 at $5.49 an hour. |
| OpenAI-compatible | Yes | Yes. Base URL https://api.together.ai/v1, compatible with the OpenAI REST API and SDKs for chat, completions, vision, images, speech and embeddings. |
| Anthropic-compatible | Yes, /v1/messages (verified with Claude Code) | Not documented as a native endpoint. Their docs run Claude Code on Together models through Together Link, a hosted gateway, or through a LiteLLM proxy you run yourself. |
| Uncensored / abliterated | Five uncensored models, the Xploded line: GLM 5.3, GLM 5.3 Flash, DeepSeek V4.1 Flash, MiMo V2.6 Flash and Qwen3.8 27B. Every one calls tools. | No uncensored, abliterated or roleplay finetunes in the catalogue, which carries official releases such as GLM-5.3, Kimi K3, DeepSeek V4 and Qwen3.8. The terms prohibit content that is obscene or constitutes pornography. |
| Prompt retention | No prompt or completion stored; model, tokens and cost kept 90 days; signed prompt-hash receipt | Privacy policy: no data collected from you is used to train their models without explicit opt-in. Prompt storage is a setting, and turning it off enables Zero Data Retention, under which prompts and outputs are not stored, retained or used for training. |
| Per-key spend caps | Lifetime and monthly per-key caps, enforced by reserving the worst case and clamping max_tokens | One balance or usage limit for the whole organisation, which blocks usage once hit. Their docs say individual API keys cannot be capped or rate-limited; limits apply at the organisation level. |
Where Together AI is the better choice
- A far larger catalogue of 200+ open models (GLM-5.3, Kimi K3, DeepSeek V4, Qwen3.8, Gemma 4), plus fine-tuning, batch inference, dedicated endpoints and GPU clusters on their own infrastructure. MagmaRouter offers none of these.
- Lower per-token prices for mainstream open models, and a documented Zero Data Retention switch plus a no-training-without-opt-in policy, which suits teams that do not need uncensored output.
Where MagmaRouter is different
- MagmaRouter sells a short line of uncensored Xploded models (GLM 5.3 Xploded, GLM 5.3 Flash Xploded, DeepSeek V4.1 Flash Xploded, MiMo V2.6 Flash Xploded, Qwen3.8 27B Xploded), all with tool calling, and serves Claude Code from its own Anthropic-compatible /v1/messages endpoint.
- MagmaRouter puts hard spend caps on each key, lifetime and monthly, enforced by reserving the worst-case cost and clamping max_tokens before the call. It is pay as you go from a $5 top-up with no subscription.
The uncensored models MagmaRouter sells
If you are comparing Together AI because you want an uncensored or abliterated model API, these are the models on MagmaRouter: GLM 5.3 uncensored, GLM 5.3 Flash uncensored, DeepSeek V4.1 Flash uncensored, MiMo V2.6 Flash abliterated, Qwen3.8 27B uncensored and Qwen3.8 27B Fast uncensored. Every one calls tools, works in Claude Code through the Anthropic endpoint, and is priced per million tokens as listed on its model page.
| Model | What it is | Context | Price in / out | Supports |
|---|---|---|---|---|
| GLM 5.3 Xploded glm-5.3-xploded | GLM 5.3 uncensored API | 1M | $2.00 / $5.00 | tools, reasoning |
| GLM 5.3 Flash Xploded glm-5.3-flash-xploded | GLM 5.3 Flash uncensored API | 1M | $0.30 / $1.00 | tools, vision, reasoning |
| DeepSeek V4.1 Flash Xploded deepseek-v4.1-flash-xploded | DeepSeek V4.1 Flash uncensored API | 1M | $0.30 / $1.50 | tools, reasoning |
| MiMo V2.6 Flash Xploded mimo-v2.6-flash-xploded | MiMo V2.6 Flash abliterated API | 1M | $1.00 / $2.00 | tools, vision, reasoning |
| Qwen3.8 27B Xploded qwen3.8-27b-xploded | Qwen3.8 27B uncensored API | 1M | $1.50 / $5.00 | tools, vision, reasoning |
| Qwen3.8 27B Fast Xploded qwen3.8-27b-fast-xploded | Qwen3.8 27B Fast uncensored API | 524K | $1.50 / $5.00 | tools, vision, reasoning |
FAQ
Does Together AI host uncensored models?
None are listed in its catalogue, which carries official releases from labs such as Zhipu, Moonshot, DeepSeek and Qwen, and its terms prohibit obscene or pornographic content. MagmaRouter's Xploded line is built for uncensored output.
Can I set a spend limit per API key on Together AI?
No. Together's docs say usage limits apply at the organisation level and individual keys cannot be capped. MagmaRouter enforces lifetime and monthly caps on each key.
Is Together AI cheaper than MagmaRouter?
For mainstream open models, usually yes: Together prices at its own rates on its own GPUs, while MagmaRouter is priced for its uncensored line and per-key caps. Together does not sell uncensored variants, so the two are not like for like.
Which uncensored models does MagmaRouter sell?
MagmaRouter sells GLM 5.3 Xploded, GLM 5.3 Flash Xploded, DeepSeek V4.1 Flash Xploded, MiMo V2.6 Flash Xploded, Qwen3.8 27B Xploded, Qwen3.8 27B Fast Xploded. The table above has the price and context for each.
Are these models abliterated or uncensored?
MiMo V2.6 Flash Xploded is abliterated: the refusal direction was removed from the weights. GLM 5.3 Xploded, GLM 5.3 Flash Xploded, DeepSeek V4.1 Flash Xploded, Qwen3.8 27B Xploded, Qwen3.8 27B Fast Xploded are uncensored: the refusal was trained back out. The abliterated vs uncensored page explains the difference.
Which one answers fastest?
MiMo V2.6 Flash Xploded usually starts answering in one to two seconds. The GLM 5.3 Flash model thinks before it answers, so its first word can take much longer.