Model catalog
Every serverless model with published per-1M-token rates — expand a row for capabilities, specs and live regional health.
Straight from the public /models_list feed, served identically from the au, eu and us regional endpoints, refreshed hourly.
| Model | Tier | Context | Input $/1M | Output $/1M | Cached input $/1M | Regions |
|---|---|---|---|---|---|---|
| DeepSeek V3.2Details → | extraLarge | 164K | $1.75 | $5.50 | — | AUEUUS |
| DeepSeek V4 Flash OfficialDetails → | medium | 450K | $0.42 | $1.51 | — | AUEUUS |
| DeepSeek V4.1 FlashDetails → | medium | 1.0M | $0.42 | $1.51 | $0.01 | EUUS |
| DeepSeek V4.1 Flash PriorityDetails → | medium | 1.0M | $0.63 | $2.27 | $0.01 | Global |
| Gemma 3 12BDetails → | small | 131K | $0.17 | $0.51 | — | AU |
| Gemma 3 27BDetails → | medium | 131K | $0.42 | $1.51 | — | AU |
| Gemma 4 26B A4BDetails → | small | 262K | $0.17 | $0.51 | — | AUEUUS |
| Gemma 4 31BDetails → | medium | 131K | $0.42 | $1.51 | — | AUEUUS |
| GLM 4.7Details → | extraLarge | 203K | $1.56 | $4.68 | — | EUUS |
| GLM 5Details → | extraLarge | 131K | $1.56 | $4.68 | — | AUEUUS |
| GLM 5.1Details → | extraLarge | 203K | $1.72 | $5.50 | — | EUUS |
| GLM 5.2Details → | extraLarge | 800K | $1.75 | $5.50 | $0.33 | AUEUUS |
| GLM 5.3Details → | extraLarge | 800K | $1.75 | $5.50 | $0.33 | AUEUUS |
| GLM 5.3 FastDetails → | extraLarge | 800K | $2.63 | $8.25 | $0.49 | Global |
| GLM 5.3 FlashDetails → | medium | 1.0M | $0.42 | $1.51 | — | EUUS |
| GLM 5.3 Flash PriorityDetails → | medium | 1.0M | $0.63 | $2.27 | $0.13 | Global |
| GLM 5.3 PriorityDetails → | extraLarge | 800K | $2.34 | $7.02 | $0.43 | Global |
| GPT OSS 120BDetails → | large | 131K | $0.84 | $2.10 | — | AUEUUS |
| GPT OSS 20BDetails → | medium | 131K | $0.42 | $1.51 | — | AUEUUS |
| InklingDetails → | extraLarge | 800K | $1.56 | $5.06 | $0.27 | EUUS |
| Kimi K2.5Details → | large | 262K | $0.84 | $2.10 | — | AUEUUS |
| Kimi K2.5 ThinkingDetails → | extraLarge | 262K | $1.56 | $5.63 | — | AUEUUS |
| Kimi K2.6Details → | extraLarge | 262K | $1.56 | $5.63 | — | EUUS |
| Kimi K2.7 CodeDetails → | extraLarge | 262K | $1.56 | $5.00 | — | EUUS |
| Kimi K3Details → | ultra | 800K | $4.20 | $21.00 | $0.42 | EUUS |
| Kimi K3 FastDetails → | ultra | 800K | $6.30 | $31.50 | $0.63 | Global |
| Kimi K3 PriorityDetails → | ultra | 800K | $6.30 | $31.50 | $0.63 | Global |
| Llama 3.1 8bDetails → | small | 131K | $0.17 | $0.51 | — | AUEUUS |
| Llama 4 MaverickDetails → | extraLarge | 131K | $1.56 | $4.68 | — | AU |
| MiniMax 2.5Details → | large | 120K | $0.84 | $2.10 | — | AUEUUS |
| MiniMax 2.7Details → | large | 180K | $0.84 | $2.10 | — | AUEUUS |
| MiniMax 2.7 PriorityDetails → | large | 180K | $1.26 | $3.15 | — | Global |
| Minimax 3Details → | medium | 180K | $0.42 | $1.51 | $0.08 | EUUS |
| MiniMax M2.5Details → | large | 1M | $0.84 | $2.10 | — | AUEUUS |
| NVIDIA Nemotron Nano 12B v2Details → | small | 131K | $0.17 | $0.51 | — | AU |
| NVIDIA Nemotron Super 3 120BDetails → | extraLarge | 164K | $1.56 | $4.68 | — | AU |
| Qwen 3 32BDetails → | medium | 33K | $0.42 | $1.51 | — | AU |
| Qwen 3.6 27BDetails → | small | 131K | $0.17 | $0.51 | — | AUEUUS |
| Qwen3 Coder 30B A3BDetails → | medium | 262K | $0.42 | $1.51 | — | AU |
| Qwen3 Coder NextDetails → | large | 262K | $0.84 | $2.10 | — | AUEUUS |
| Qwen3 VL 235B A22BDetails → | extraLarge | 262K | $1.56 | $4.68 | — | AUEUUS |
| Qwen3.8 27BDetails → | small | 262K | $0.19 | $2.34 | — | AUEUUS |
| Qwen3.8 MaxDetails → | ultra | 1M | $2.60 | $7.80 | $0.33 | EUUS |
| TF MysticaTF modelDetails → | extraLarge | 800K | $2.34 | $7.02 | — | AUEUUS |
| TF Mystica ThinkingTF modelDetails → | extraLarge | 800K | $2.34 | $7.02 | — | AUEUUS |
| TF SorcererTF modelDetails → | medium | 800K | $0.63 | $2.27 | $0.16 | AUEUUS |
| TF Sorcerer ThinkingTF modelDetails → | large | 800K | $1.26 | $3.15 | $0.32 | AUEUUS |
Model catalog questions, answered
Rates, residency and regional health — the five questions every team asks before picking a model from this table.
Where do these rates come from?
Straight from the public /models_list feed — the same feed our own platform bills from — refreshed hourly and served identically by the au, eu and us regional endpoints.
There is no negotiated rate card behind this table: the per-1M-token price you see quoted is what you are billed.
Which models keep my data in Australia?
Call the AU endpoint: every regional deployment runs smart routing that keeps requests inside the region you call — AU requests are processed in Australia. The same holds for the EU and US endpoints. The AU, EU and US chips show which regional endpoints serve each model.
What do the region chips mean?
They show which regional endpoints serve a model: AU means it is served on the Australian endpoint, EU in Europe and US in the United States — and a model can be served in several regions at once. Whichever regional endpoint you call, smart routing keeps processing inside that jurisdiction.
Global is different: it means the model is not served from any regional endpoint today and is routed wherever capacity is available. Filter by AU, EU or US to list only the models served on that regional endpoint.
What does the health score in the expanded row mean?
Each expanded row shows the live health score for each region, pulled from the same live feed and refreshed hourly.
Figures reflect the latest telemetry reported by each region's data centres and update automatically as the feed refreshes.
How do I start using a model from the catalog?
Grab an API key and call the OpenAI-compatible endpoint for your preferred region — every model in the table is reachable with the same key.
Swapping models is a single string change in your code, there are no seats or licences, and platform access starts from $5 of pay-per-use credits.