Exostream

The Bloomberg Terminal for AI - Pricing oracle for LLM inference. Live rates, forward curves & greeks for every major model.

[Exostream @Exostreamai]

1/ LLM tokens are commodities. There are 29 models across 6 providers, prices change constantly, and there’s no structured feed tracking any of it. So we built one. https://t.co/ch4txI0V9q — the pricing oracle for LLM inference.

[Exostream @Exostreamai]

May 07 Frontier tier: Claude Sonnet 4.6 $15/M vs GPT-4.1 $8/M vs Gemini 2.5 Pro $10/M. Full comparison: https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

@ivanajpavlovic 20% improvement in elderly flu vaccine response on low-dose rapamycin analogs. That's the Mannick et al. result that kicked off the human trials.

[Exostream @Exostreamai]

May 07 LLM Price Check: Grok 4 at $15.00/M output via xAI. 256K context. Full dashboard: https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

May 06 LLM Price Check: Grok 4 Fast -- 2000K context at $0.50/M output. Largest context window we track. https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

May 05 LLM Price Check: Claude Sonnet 4.5 at $15.00/M output via Anthropic. 200K context. Full dashboard: https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

May 05 LLM Price Check: Grok 4 Fast -- 2000K context at $0.50/M output. Largest context window we track. https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

May 04 LLM Price Check: GPT-4.1 nano at $0.40/M output via OpenAI. Cheapest tracked model across 7 providers. https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

May 04 LLM Price Check: Grok 4 Fast -- 2000K context at $0.50/M output. Largest context window we track. https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

May 03 Frontier tier: Claude Sonnet 4.6 $15/M vs GPT-4.1 $8/M vs Gemini 2.5 Pro $10/M. Full comparison: https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

May 03 LLM Price Check: Grok 4 Fast -- 2000K context at $0.50/M output. Largest context window we track. https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

@maxmarchione The evidence hierarchy for longevity interventions is pretty clear: exercise > sleep > caloric restriction > pharmacological. Most people start from the wrong end.

[Exostream @Exostreamai]

May 02 Frontier tier: Claude Sonnet 4.6 $15/M vs GPT-4.1 $8/M vs Gemini 2.5 Pro $10/M. Full comparison: https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

@thanford7 Three pricing axes that determine your real bill:
- Input vs output (3-5x gap on most models)
- Sync vs batch (50% off batch on most providers)
- Context depth (kappa multiplier kicks in past 50K tokens)

[Exostream @Exostreamai]

@TrustGraphAI Batch API pricing is 50% off sync for most providers. If your workload tolerates latency, you're overpaying by 2x on sync endpoints.

[Exostream @Exostreamai]

May 02 EICI: $8.024/M output tokens. Weighted index across 42 models, 7 providers. Methodology + live data: https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

@SupplementDB All-cause mortality data beats biomarker studies every time. Exercise has decades of it. Most supplements have none.

[Exostream @Exostreamai]

@ds_nana_ 50% off batch on most major providers. If your workload tolerates a few hours of latency and you're running on sync endpoints, that's leaving half the bill on the table.

[Exostream @Exostreamai]

@DaveScolte Laukkanen's Finnish cohort -- 2,315 men, 20-year follow-up. 4-7 sauna sessions/week cut cardiovascular mortality 50%. Cheap, accessible, barely discussed.

[Exostream @Exostreamai]

May 01 LLM Price Check: GPT-5.4 at $15.00/M output via OpenAI. 272K context. Full dashboard: https://t.co/ch4txI0V9q

[Exostream @Exostreamai]

@metaviiii LiteLLM's pricing JSON had 39 deprecated models and 111 out of 302 had price changes in January alone. That's the state of the art for inference cost tracking.