AGPL-3.0-or-later · forever.

the provider catalog

Google Gemini

The Gemini API: a fast flash tier and long context, with keys managed in Google AI Studio. One provider/model line moves a workflow here or away again: the same file runs on every provider the catalog names, and swapping is a one-word diff. Machines read the catalog; the binary answers nika catalog.

  • 2pinned modelsaliases pin exact upstream ids
  • cloudfamilybring your own key; it rides an env var and stays yours
  • GEMINI_API_KEYthe key rides herean env var, never a config file
  • geminiwire dialectthe adapter speaks it

the wire

cloud · bring your own key; it rides an env var and stays yours

A workflow names this provider in the model line (model: gemini/gemini-2.5-flash), and the engine speaks gemini on the wire. The key rides GEMINI_API_KEY in the environment, never a config file, never the workflow. The provider id also answers to google.

default
gemini-2.5-flash
cheap
gemini-2.0-flash
dialect
gemini
tags
code-execution · frontier · function-calling · long-context · multimodal · streaming · vision · web-search

the models

2 pins

Aliases are what the file may write; the engine pins each one to an exact upstream id, so a run stays reproducible after a vendor rotates names.

  • flash-25gemini-2.5-flashctx 1.0m · vision · json:schema
  • flash-20gemini-2.0-flashctx 1.0m · vision

in the released catalog

2 rooms

The wire models this provider serves in the released binary's catalog · each one is a room with its seats, exact prices and measured energy.

the whole register: /catalog/models · what the released binary knows, vendored at the engine pin.

in a real file

a complete workflow · shown whole
gemini.nika.yaml
nika: v1workflow:  id: flash-digest  description: "a fast, steady digest — GEMINI_API_KEY in the environment, nothing else"tasks:  digest:    infer:      model: gemini/gemini-2.5-flash      max_tokens: 150      temperature: 0.2      prompt: |        Digest to three bullets: the run is a DAG, the DAG comes        from references, references are checked before anything runs.outputs:  bullets: ${{ tasks.digest.output }}

check-green against the released engine · 1 task · 1 wave · no closed ceiling: an unpriced model is unpriced, never free · the drift gate re-proves this copy on every test run · open it in the playground →

get set up

keys at aistudio.google.com
  1. 1Create a key in AI Studio (keys are project-scoped).
  2. 2The key rides the environment, never the file export GEMINI_API_KEY=…

Pick it for quick digests and long inputs on a generous free tier. Then prove the seat: nika doctor names every provider this binary can reach.

swap it

10 more cloud seats

Moving off Google Gemini is editing one line. The rest of the file (tasks, permits, outputs, the trace) does not change. The register orders the catalog local first; beyond it the engine embeds 21 more OpenAI-compatible endpoints.

cross-references

the words that meet it

Try it in the playground, or install and go local first. Read the spec →