Google / PaLM 2 for Chat
Historical PaLM 2 hosted chat release, with original version identifiers and dated evidence.
Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
SDK explicitly describes these RLHF allowlist models as supporting both online and batch prediction.
Exact identifiers appear in the SDK RLHF allowlist; historical support only.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
Exact historical model-version evidence has not been collected; no current replacement value is inferred.
No exact parameter count or underlying checkpoint equivalence is established.
Retained SDK example, not a live service check.
These are training sequence settings, not model inference context or output limits.
Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.
chat-bison@001Endpoint pricing has not been collected. Check the provider for applicable rates.
Retention, training use, regional processing and enterprise terms have not been summarized for this endpoint.
Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.
The system processes eligible input and retains reusable state.
Reuse depends on matching input, session state and the cache policy.
Reusing state can reduce repeated work. Any charges follow the billing rules below.
chat-bison@001No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified historical caching policy; contemporary Gemini caching does not apply automatically.
No verified cache rate for this endpoint. An empty rate is not a zero charge.
Caching reviewed 14 Oct 2026No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.
Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.
Google documents this proprietary hosted model or version through original release notes and SDK identifiers. No public weights are supplied in the reviewed evidence. Exact architecture, model size and equivalence across versions are not inferred from the client SDK license.
Classification evidence (opens in a new tab)