Topic
Software layers, such as LiteLLM and OpenRouter, that let applications call multiple LLM providers through one API, handling routing, failover, and billing.
LiteLLM 1.83.9 fixes two Check Point-reported flaws, CVE-2026-35029 and CVE-2026-59823, that affect every older version of the gateway. A DEV Community patch guide pairs the upgrade with key rotation and an audit of what the gateway held.
Reality
agentgateway v1.6.0 priced a Claude Sonnet 5 request at $0.000156 from its built-in catalog, with no hand-kept modelCatalog in the config. The test covered one Anthropic model, so retiring rate tables is safe only for models the catalog provably knows.
Reality
LiteLLM's standalone Rust path added 0.7 ms of p99 latency against 257.7 ms for its Python proxy in July 22 benchmark artifacts reviewed on dev.to. Logging, persistence and spend tracking were off for the run, so the gap measures forwarding alone.
Reality
LiteLLM launched Lens on September 30, a tool that uses AI agents to find recurring failures across agent traces sent through its model gateway. Customers host the analyzer and its databases, and the 200,000-trace volume CTO Ishaan Jaffer cites is a future target Lens has not been measured against.
Reality
One developer's no-key test of four LLM gateways found model lists out of step with vendor docs, the author's own falling from 477 to 199 models in a week. Any model ID read from these lists is a dated observation and should be snapshotted before it goes into code.
Reality
Chinese AI models handled 57% to 67% of OpenRouter's tokens in mid-September, up from 6% to 13% in February, according to data shared with CNBC. Because the models win on price, most of the tokens can still be a minority of the money.
Reality
Model identifiers are untyped strings, so a one-character slip on a multi-vendor gateway picks up a different cache-read rate with no error and nothing in the logs. A dev.to post turns the pricing page into a CI check.
Reality
A dev.to writeup pits Bifrost against LiteLLM on a shared four-core VPS, using the harness the Bifrost team maintains. The widest gap it reports traces back to a worker default the official LiteLLM image leaves unset.
Reality
The Responses and Messages dialects specify tool declarations, streaming and reasoning knobs. A Java 21 library published to Maven Central at 0.18.0 argues the unspecified half is where retrieval, injection and veto actually live, and offers to host it in your own process.
Reality
The MIT-licensed Node gateway wraps thirteen cloud providers and local servers in one call shape, and the dollar figure it returns for each call comes out of a catalog your own team transcribes from provider pricing pages.
Reality
The eleven-check suite comes from a gateway maintainer who wrote it to be pointed at their own endpoint as well as everyone else's, and the checks that matter most are the ones where a divergence still answers with success.
Reality
A dev.to walkthrough replaces developer-held provider keys with a gateway that tags every call with org, team and user. The tagging is server-side, which is the right call, and the dashboards behind it refresh every 15 minutes.
Reality
A dev.to practitioner argues the deciding test is a per-tenant usage ledger and a JSON contract that holds on a fixed labeled eval set. Price enters last, if at all.
Reality