Arrow 2 Telos
A model for complex SVG creation with a larger context window.
Arrow 2 Telos is designed for complex SVG creation. Its larger context window provides more room for instructions and source material. Use arrow-2-telos as the API model ID.
Availability
arrow-2-telos is generally available. List it with GET /v1/models using either a Production or a Test key. Your key must also authorize the model and operation; do not assume a key may call it from this documentation page alone.
The public API supports SVG generation, vectorization, editing, and animation when enabled for your key. Use Responses only when the returned supported_operations includes open_responses. Check the matching operation before calling a native SVG endpoint.
Model details
| Property | Value |
|---|---|
| API model ID | arrow-2-telos |
| Input modalities | Text, image, SVG |
| Output modality | SVG |
| Context window | 1,050,000 tokens |
| Maximum output | 65,536 tokens |
| Billing kind | token_usage |
Supported request fields vary by endpoint.
On native SVG generation and vectorization, omit temperature, top_p, and presence_penalty. Telos rejects any explicit value for these fields, including 1, 0, or null, with HTTP 400. This native endpoint rule is separate from the existing Responses sampling restrictions.
Pricing
| Input | Cached input | Cache write | Output |
|---|---|---|---|
| $6.00 | $0.60 | $7.50 | $30.00 |
All prices are USD per 1 million tokens. Telos rates are 50% higher than Arrow 2 in every category. The same token counts and cache mix cost 1.5× as much, but actual request totals depend on each model’s measured usage. Read current rates from billing.rates and see Pricing for cache accounting and a worked comparison.
Start building
Follow the Quickstart, using arrow-2-telos only after your catalog confirms access and the required operation. For an existing integration, see Migrate to Responses.