Compare models
Choose two or three models, including exact thinking and fast versions. Copy the URL to share your selection.
| Comparison | GPT-6 Astraopenai/gpt-6-astra | Choose model 2 |
|---|---|---|
| Type | Chat | — |
| TokenBaazar price · 50% off | ₹480.00 in / ₹2400.00 out per 1M tokens | — |
| Market price | ₹960.00 in / ₹4800.00 out per 1M tokens | — |
| Context window (provider) | 1,050,000 tokens | — |
| Input token limit (provider) | Within the context window | — |
| Maximum output | 128,000 tokens | — |
| Knowledge cutoff | Apr 30, 2026 | — |
| Vendor input modalities | Text, Image | — |
| Vendor output modalities | Text | — |
| TokenBaazar endpoints | POST /v1/chat/completions | — |
| TokenBaazar input | Text, Image/PDF content through the chat bridge | — |
| TokenBaazar output | Text | — |
| Benchmark evidence | 6 source-linked results | — |
| FrontierMath Tier 4 (v2) | 97.6% Third-party reproduction of OpenAI’s announcement table; reported maximum effort unless noted, not a TokenBazaar run. Direct launch page could not be fetched during this audit. Vellum · GPT-6 Astra launch-table analysis | — |
| ARC-AGI-3 | 99.9% Semi-private ARC-AGI-3, Provider Adapter harness, high reasoning: 99.9%. Standard harness at max reasoning instead reports 62.7%; opaque state reuse and compaction materially change the result. ARC Prize · GPT-6 Astra evaluated results | — |
| ExploitBench | 100% Third-party reproduction of OpenAI’s announcement table; reported maximum effort unless noted, not a TokenBazaar run. Direct launch page could not be fetched during this audit. Evaluated without production safeguards. Vellum · GPT-6 Astra launch-table analysis | — |
| Terminal-Bench-Science 0.1 | 64.6% Third-party reproduction of OpenAI’s announcement table; reported maximum effort unless noted, not a TokenBazaar run. Direct launch page could not be fetched during this audit. Vellum · GPT-6 Astra launch-table analysis | — |
| FrontierCode v1.1 (Extended) | 60.25% Cognition FrontierCode 1.1; GPT-6 Astra; medium effort; codex harness; Extended subset (150 tasks). Published weighted rubric score (new_score × 100), not the all-or-nothing pass rate; solution-source violations score zero. Source field new_score=0.6025. Publisher evaluation, not a TokenBazaar measurement; provider effort configuration is not a guarantee of the API's fixed thinking budget. Cognition · FrontierCode 1.1 original leaderboard | — |
| FrontierCode v1.1 (Main) | 48.83% Cognition FrontierCode 1.1; GPT-6 Astra; medium effort; codex harness; Main subset (100 tasks). Published weighted rubric score (new_score × 100), not the all-or-nothing pass rate; solution-source violations score zero. Source field new_score=0.4883. Publisher evaluation, not a TokenBazaar measurement; provider effort configuration is not a guarantee of the API's fixed thinking budget. Cognition · FrontierCode 1.1 original leaderboard | — |
No overall score, winner or cost-versus-score frontier is calculated without comparable, attributable benchmark evidence. Different test configurations appear separately. Vendor capabilities are distinct from TokenBaazar API support.
— means no published value or no matching capability. Provider token limits are not tested gateway guarantees; TokenBaazar media controls and response formats are listed separately.
GPT-6 Astra
Sources and verification
Provider specifications describe the upstream model, not a guarantee of every feature through TokenBazaar. Prices come from the enabled catalog; benchmark results belong to the exact tested model.
- Provider model specifications
Vendor-reported
Last checked - Vellum · GPT-6 Astra launch-table analysis
Third-party report
Last checked - ARC Prize · GPT-6 Astra evaluated results
Third-party report
Last checked - Cognition · FrontierCode 1.1 original leaderboard
Third-party report
Last checked
Research notes (1)
- Published provider limits; maximum-capacity requests have not been independently exercised through TokenBaazar. Thinking-level variants use the same provider model; reasoning and answer tokens share the output budget.