recommend deployment

$0.25 USDC: recommend what to buy or run from bounded StackBench fit, performance, confidence, evidence, caveats, and economics only when supported.

liveverifiedunknown~5s typical

mcpprobedcomputesagestackbench

$0.0020

per call

plus $0.001 platform fee per run

Open MCP server; Cracked bills routing only

Provider Upstream docs

$Set up https://cracked-ai-pearl.vercel.app/SKILL.md, then use Cracked to run mcp-com-computesage-stackbench /recommend_deployment for me.

Input

Field	Type	Description
model*	string
hardware_system	string
mode	string (configuration \| hardware_experimental)	default "configuration"
quantization	string
engine	string
engine_version	string
prompt_tokens*	integer
decode_context_depth	string
batch_size	string
ubatch_size	string
concurrency	integer	default 1
flash_attention	string
candidate_kv_formats	string
minimum_prefill_tok_s	string
minimum_decode_tok_s	string
maximum_vram_bytes	string
maximum_power_watts	string
objective	string (balanced \| maximize_decode \| maximize_prefill \| minimize_vram \| minimize_power \| minimize_capex)	default "balanced"
evidence_policy	string (verified_only \| include_public \| all)	default "include_public"
constraint_policy	string (point_estimate \| conservative)	default "point_estimate"

Call it

curl

curl https://cracked-ai-pearl.vercel.app/v1/run \
  -H "Authorization: Bearer ck_live_..." -H "content-type: application/json" \
  -d '{"provider":"mcp-com-computesage-stackbench","endpoint":"/recommend_deployment","input":{"model":"...","mode":"configuration","prompt_tokens":1,"concurrency":1,"objective":"balanced","evidence_policy":"include_public","constraint_policy":"point_estimate"}}'

cli

npx cracked-ai run -p mcp-com-computesage-stackbench -e /recommend_deployment -i '{"model":"...","mode":"configuration","prompt_tokens":1,"concurrency":1,"objective":"balanced","evidence_policy":"include_public","constraint_policy":"point_estimate"}'

mcp

run_tool({ provider: "mcp-com-computesage-stackbench", endpoint: "/recommend_deployment", input: {"model":"...","mode":"configuration","prompt_tokens":1,"concurrency":1,"objective":"balanced","evidence_policy":"include_public","constraint_policy":"point_estimate"} })

Try it

Runs against your signed-in workspace balance. Sign in if you have not.