Fastino GLiDE (hosted API, 22 Oct 2026)

Composite (secondary)

76.6 · Not ranked on this board

Cost and measurement conditions

$0.023 per 1,000 decisions (not published)

Operator's own published standard usage tariff: USD 0.04 per 1M INPUT tokens, output not billed (docs.fastino.ai/pricing, GLiDE row; page read 22 Oct 2026 20:34 UTC, saved copy sha256 22c0758f). Counted under interpretation I-1 (a manufacturer's standard, non-promotional list price counts from day 1) by the frozen pricing module pricing_v15 (sha256 aa2afcd5a58d0c4cc0e3d6867d590c70b6ac58e971410c27faf9628f97bcf8b7), which carries one added launch-list-price key for this row and nothing else. Applied to this row's own provider-reported input tokens on the answered rows; the 17 requests that returned an error are priced at the API's 65,536-token prompt ceiling, the documented context estimate. COMMON COST BASIS (v1.6.30): cost per 1,000 decisions is computed on the common items of this draw, leaving out the 15 very long items that count for no row; this row scored 1,497 items, so the basis here is 1,482 items. This row: USD 0.023065 per 1,000 decisions (0.357x the official cost cap USD 0.064595), against USD 0.049101 over all 1,497 scored items. The older measurement of this vendor's GLiDE on 6 Oct 2026 (key fastino-glide) was priced at the then tariff of USD 0.15 per 1M input tokens and is unchanged.

p50 latency: 0.4 s

none (API)

Vendor-hosted api.fastino.ai /v1/systemone API, model `fastino/GLiDE` (responses report glide_v2), evaluator key, every request with store:false; latency measured from Sandy (Germany) on every item without API adjustment. Full S+P draw of v1.6-regular-20261012 minus the 3 custody-excluded items (1,497 requests, one pass, no retries) through the unchanged run_api_full_v161.py; the full-set exposure gate logged the 1,197 sealed items to the provider before the first request. Measured 22 Oct 2026 by the evaluator; the vendor did not take part in the run.