A model-delegation router for Claude Code and Codex: lookups, checks and test runs go to cheaper models, and the savings show up in your statusline.
Same 12 requests, two terminals: $6.68 → $1.83 (−72.6%) with Sprag.
Example requests priced at public list rates. Full video (41s)
Sprag routes each piece of work to the cheapest model that has never got it
wrong. Claude Code and Codex cannot switch models per prompt, so Sprag does it
the one way that works: rules in the main session hand lookups, status checks,
build and test runs and batch edits to sub-agents pinned to a cheaper model. The
rules come from your own session logs, you approve each one, and every delegated
run writes its price difference to a ledger. npm i -g sprag-cli sets it all
up; nothing to configure per project.
Two more things ride along. Documents reach the context as compact Markdown (a 540k-token deck becomes 23k tokens), and the statusline shows what each session spent and saved, with a warning chip in front when the cache, the context or a rate limit needs attention.
The name is the sprag clutch: forward motion passes, backspin locks. Repeated failures in your logs become rules loaded at session start, so the agent keeps working and only the regressions get caught.
Session analysis runs locally on the logs Claude Code and Codex already write. No extra model calls or API key are needed for analysis. Optional Codex gateway budget and price lookups contact your configured LiteLLM endpoint using its existing credentials. The CLI requires Node.js 20+ and one TOML parser dependency.
npm i -g sprag-cli # Claude Code, plus Codex when it is installed
If your package manager skips install scripts (pnpm, bun), run sprag install once after installing. The old claude-token-saver package stopped updating at 3.55.0, so move to sprag-cli.
After you install
- Run
npm i -g sprag-cli. Hooks, the statusline and the starter rules are set up for every project; there is nothing to configure per project. If you already have a statusline, the install asks whether to show both, replace yours, or keep yours. - Open Claude Code in any project and work as usual.
- Watch the statusline. It shows what the session spent and saved, and a warning chip leads the line when the cache, the context or a rate limit needs attention:
🔀 Routing saved $188 | opus→sonnet ×67 $93.4 · fable→sonnet ×26 $83.7 · opus→haiku ×25 $7.41 · fable→haiku ×6 $3.73
🧠 Cache hit 98.8% · ⏳ Cache expires 59:46 · 📦 Ctx 47% of 1M
Sprag never swaps the model you picked. Analysis reads your local session
logs only, with no extra model calls and no API key. sprag uninstall removes
everything it added and restores a statusline it merged with.
Using Codex? See Codex below.
How it works
Everything on the left is something your session already produces. Nothing is sent anywhere to read it. The tier decision at the bottom is the whole routing argument in three tests: where the answer lives, whether a mistake surfaces, and what error rate the rule was measured at.
Sprag does not swap the model you asked for. It reads what you have already run, finds the kinds of task that never went wrong on a cheaper tier, and spawns a sub-agent for exactly those. The main agent reviews what comes back before anything counts as done, and the work returns to it the moment a rule stops holding.
Measured results
| Metric | Result | Evidence | |
|---|---|---|---|
| Public benchmark accuracy | 59.1% vs 57.9% best single model | ▰▰▰▰▰▰▰▰▰▰▰▰ |
benchmark |
| Cost for the same queries | $268 vs gpt-5 $388, gemini-2.5-pro $734 | ▰▰▰▰▰▱▱▱▱▱▱▱ |
benchmark |
| Tokens for the same documents | −95.9%, 2,011,178 → 82,209 over 12 files | ▰▱▱▱▱▱▱▱▱▱▱▱ |
doc2md |
| Delegation savings, from the ledger | $188.27 over 124 delegated runs | ▰▰▰▰▰▰▰▰▰▰▰▱ |
sprag route-scan savings |
| Cheap tier on the same prompts | 24% of the cost, one answer worse out of eight | ▰▰▰▱▱▱▱▱▱▱▱▱ |
A/B run |
The first two rows answer the objection that comes up first: that moving work to a cheaper model has to cost you accuracy. It did not. The same criteria, wired up as a router, scored 1.2 points above the best single model at a third of the cost. Nothing is best at everything, which is why choosing per task beats choosing once.
Measured offline on LLMRouterBench (Findings of ACL 2026): 11,696 instances, 13 models whose answers, scores and costs ship with the dataset. No model was called and nothing was spent. The router chose; the numbers were looked up. Run it again and you get the same figures. It validates the criteria, not sprag's own routing: the dataset has no Claude models, so deepseek-v3, qwen3-235b and gpt-5 stood in for cheap, mid and flagship.
The last three rows are ledgers on this machine. They grow as it runs. The routing figures were read on 2026-10-09, the document figures on 2026-09-18. Model names are the ones each run actually used.
Current, on one machine, last 30 days
🔀 Routing saved $188.27 total (last 7d $13.31 · last 30d $150.64)
fable-5 → sonnet-5 24 runs $81.19 ▰▰▰▰▰▰▰▰▰▰▰▰
opus-5 → sonnet-5 54 runs $68.85 ▰▰▰▰▰▰▰▰▰▰▱▱
opus-5 → sonnet-5.5 13 runs $24.58 ▰▰▰▰▱▱▱▱▱▱▱▱
opus-5 → haiku-4.5 25 runs $7.41 ▰▱▱▱▱▱▱▱▱▱▱▱
fable-5 → haiku-4.5 5 runs $3.22 ▰▱▱▱▱▱▱▱▱▱▱▱
fable-5 → sonnet-5.5 2 runs $2.51 ▰▱▱▱▱▱▱▱▱▱▱▱
fable-5 → haiku-5.5 1 run $0.51 ▰▱▱▱▱▱▱▱▱▱▱▱
| Also measured | Over 30 days | |
|---|---|---|
| Documents converted | 12 files, 2,011,178 → 82,209 tokens | ▰▱▱▱▱▱▱▱▱▱▱▱ |
| Cache hit rate | 92.2% | ▰▰▰▰▰▰▰▰▰▰▰▱ |
| Volume behind both | 175 sessions · 14,682 API calls · 3.81B input tokens |
Real output, from sprag route-scan savings, sprag --days 30, and the doc2md
ledger. Nothing here is a projection.
Every delegated run records its own price difference, so sprag route-scan savings traces any amount back to the rule that caused it. What the ledger deliberately leaves out is written down in the command reference.
Research behind the tier rules
The tier criteria are not a hunch. Each one traces back to published routing research, and the percentages quoted in the rules (28% vs 54% recall, a +63pp gap on chained tool work) are sprag's own offline measurement on the datasets below.
| Criterion | What the paper found | Paper |
|---|---|---|
| Delegate down, conservatively | As budgets grow, learned routers default to the most expensive model even when a cheaper one already suffices. | Lai & Ye, When Routing Collapses: On the Degenerate Convergence of LLM Routers (2026) |
| Judge difficulty by outcome | Every router on the board falls short of the oracle, mainly because it cannot tell when a smaller model is enough. Query difficulty is defined by how many of 42 models actually answered it. | Lu et al., RouterArena: An Open Platform for Comprehensive Comparison of LLM Routers (2025) |
| Thresholds from your own logs | Routing compares a predicted win rate against a cost threshold, and results are read by the share of calls sent to the strong model. sprag sets that threshold from your last 14 days at every scan. | Ong et al., RouteLLM: Learning to Route LLMs with Preference Data (2024) |
| Escalate on failure | A cascade tries the cheap model first and lets a check on the answer decide whether to go up. It matched GPT-4 with up to 98% lower cost. | Chen, Zaharia & Zou, FrugalGPT (2023) |
| Where the answer lives | Short-form factual recall measured on its own. The 28% vs 54% recall gap comes from this class of set. | Wei et al., Measuring short-form factuality in large language models (SimpleQA) (2024) |
| Tool orchestration stays up | Agents must coordinate tool calls with a user in a shared environment. The +63pp gap was measured here. | Barres et al., τ²-Bench (2025) |
| The public benchmark run | 400K instances, 21 datasets, 33 models. Several routers, commercial ones included, fail to beat a simple baseline. | Li et al., LLMRouterBench (Findings of ACL 2026) |
| Routing vs cascading | A taxonomy of when a routing decision is made, what it reads, and how it is computed. | Moslem & Kelleher, Dynamic Model Routing and Cascading for Efficient LLM Inference: A Survey (2026) |
Two criteria have no paper behind them. Batching tasks that read the same files and splitting design from code writing come from sprag's own cost accounting: on a sub-agent, loading context costs more than generating. The full reasoning, in Korean, is in Tier criteria.
Everything ships in one install, working from day one
| What it does | More | |
|---|---|---|
| ⚙️ Ratchet rules | Repeated failures become one-line rules loaded every session. Candidates are detected from your logs, and you choose project or global scope. | harness |
| 🔀 Model fitting | Log-driven delegation rules carrying the error rate they were measured at and the savings they reported, written to ratchet-model.md. |
route-scan |
| 🅷 Harness score | Five operating principles checked live. Skip the verify step and 🅷 4/5 says so before you report done. |
harness |
| 📊 Token telemetry | Cache hit rate, TTL, context size, output spikes and both rate-limit windows, in the prompt every turn. | statusline |
| 📄 doc2md | pptx, xlsx, pdf, docx and fig converted on demand, so a document costs one read instead of your whole context. | doc2md |
| 🇰🇷 Style gates | Write-time prose lint enforced by hook. Shipping today for Korean technical writing: double passives, translationese, cohesion. | style |
The two savings figures are never added together. They measure different things: the routing ledger records the price difference on delegated runs, and the conversion ledger records tokens a document did not cost.
Neither one counts what the harness and the ratchet rules save. That saving arrives as fewer round-trips rather than as cheaper ones.
The statusline in one line
🔀 Routing saved $188 | opus→sonnet ×67 $93.4 · fable→sonnet ×26 $83.7 · opus→haiku ×25 $7.41 · fable→haiku ×6 $3.73
🚨 5H ▰▰▰▰▰▰▰▰▰▰▰▱ 94% 🔄 12:36 · 🤖 Opus 5.5 · 🧠 Cache hit 98.8% · ⏳ Cache expires 59:46 · 📅 weekly ▰▰▰▰▰▱▱▱▱▱▱▱ 38% · 💵 Sep $42 · 📦 Ctx 47% of 1M
When something is wrong the warning chip leads the line: 🚨 5H/7D NN%, ⚠ Ctx 500k+, ⚠ Cache miss, ⚠ Input spike, ⚠ Output heavy, ⚠ Call surge, ⚠ Rebuild churn, ⚠ 5m TTL. Paste the chip text into Claude and the Skill names the root-cause code and the fix. Every segment is explained in the statusline reference.
Inside a JetBrains IDE terminal the chips render as single-cell glyphs instead of emoji (◉ Cache hit 98.8% · ◧ Ctx 47% of 1M). The IDE's default font has no emoji glyphs, so emoji leave characters from the previous frame on screen, like Cache expires 4:545. Label modes covers the detail, plus sprag mode narrow and mode icon-force.
The cheap tier on the same prompts
The benchmark uses public data. This asks the same question of my own work: eight
prompts from real sessions in this repository, sent to both tiers, graded against
an answer written before the run (2026-09-17, raw data in
video/out/measure2.json).
| flagship (opus-5) | cheap tier (haiku-4.5) | |
|---|---|---|
| Cost for all eight | $2.5652 | $0.6132 (23.9%) |
| Graded correct | 7 / 8 | 6 / 8 |
| Gave the same answer | 6 / 8 |
The prompt both tiers missed is the useful one. It asked which files reference
process.env without saying whether subdirectories counted, and the two read it
differently. Adding that one condition made their answers identical. The question
was ambiguous, not hard.
The remaining miss is the cheap tier leaving entries out of a file listing. A wrong listing is one you can see is wrong, which is the second criterion doing its job.
So the cheap tier is not free. On this sample one lookup in eight comes back worse, for a quarter of the price. That is why rules only send work whose failure surfaces, and never send design or diagnosis.
What the method has produced here
| Measured | Value | Sample | Window |
|---|---|---|---|
| Documents converted | 2,011,178 → 82,209 tokens | 12 files (xlsx, pptx, pdf) | to 2026-09-18 |
| Delegated runs | $188.27 saved, $1.52 per run | 124 runs | to 2026-10-09 |
| Ratchet rules accumulated | 37 (30 global, 7 per-project) | 4 repositories | 2026-05-08 to 09-16 |
The two counts grow differently. Ratchet rules come from mistakes a person made, so they keep coming. Delegation rules stop: the criteria recognize a fixed set of work types, and once each has a rule there is nothing left to promote.
Codex
Codex gets native hooks, an AGENTS.md harness, scoped ratchet rules, Korean lint, document conversion, token reports and a session-bound companion panel. npm i -g sprag-cli sets Codex up too when it is installed, with the starter rules registered globally and nothing asked (sprag install --manual asks at each step). For a Codex-only install that leaves Claude Code alone, run npm i -g sprag-cli --ignore-scripts and then sprag install --agent codex.
Commands default to Claude Code; add --agent codex to pick Codex, for example sprag --agent codex --days 7 or sprag panel --agent codex. Review and trust the new hooks in Codex's /hooks menu, because Codex requires that step. The dollar figures in this README come from Claude Code sessions, not from a Codex benchmark.
The full command list, the Claude Code and Codex comparison, accounting limits and removal are in Codex support and limits (source), or run sprag capabilities --agent codex for the support matrix.
Presets
The ratchet presets, model-routing presets and delegation agents in presets/ are also published centrally from the Sprag console. Once a day the CLI asks the Sprag server whether a newer version exists, a read-only request that carries nothing about you, and applies it only when its signature verifies against a key built into the CLI. Offline, blocked or unverified, it keeps the copy bundled with the package. A new version takes effect at the next session start, and presets you already accepted with sprag seed pick up reworded text on their own; new presets are still offered one at a time.
sprag presets status shows which copy is in use and why, sprag presets sync checks now, sprag presets show prints the effective presets, and SPRAG_NO_PRESET_SYNC=1 turns the fetch off. The request is listed in the network table of the install guide.
Day-to-day commands
| Command | What it does |
|---|---|
sprag |
Last-30-day diagnostic report |
sprag handoff |
Back work up to markdown before a cap blocks you |
sprag route-scan |
Find delegation candidates in your own history (0 LLM calls) |
sprag route-scan savings |
The routing-savings ledger |
sprag harness check |
Current 🅷 harness score |
sprag doc2md <file> |
Convert one document to Markdown |
Every subcommand, flag and the output-language setting: command reference.
Documentation
| Doc | Covers |
|---|---|
| Codex | Installation, panel, native hooks, automatic candidates, presets, and accounting limits |
| Install | What the install turns on, what stays manual, how to turn each part off |
| Statusline | Segments, warning chips, spike issue codes, update notifications |
| Commands | Full CLI surface |
| Harness | The five principles, rule promotion, compact-window pinning, seed presets |
| route-scan (한국어) | Tier verdicts, rule-file mechanics, scan triggers, subagent setup |
| Tier criteria | Why T0/T1/T2 split where they do, with sources (Korean) |
| Benchmark | The 11,696-instance LLMRouterBench run |
| doc2md (한국어) | Supported formats, savings evidence, editing documents, DRM cases |
| [Korea |
