Scout Before You Route: Predicting Task Success and Token Count from a Small LLM Prefill

Developers and consumers increasingly choose reasoning models (RMs) based on their listed API prices. However, how accurately do these prices reflect actual inference costs? We conduct the first systematic study of this question, evaluating 8 frontier RMs across 12 diverse tasks covering competition math, science QA, code generation, and multi-domain agents. We uncover the pricing reversal phenomenon: in 32% of model-pair comparisons, the model with a lower listed price actually incurs a higher total cost, with reversal magnitude reaching up to 28x. For example, Gemini 3 Flash's listed price is 80% cheaper than GPT-5.4's, yet its actual cost across all tasks is 38% higher. We build a formal cost attribution framework based on Shapley value, and leverage it to trace the dominating contributors to vast heterogeneity in thinking token consumption and number of interaction turns: on the same query, one model may use 900% more thinking tokens than another, or 10x more turns of environment interactions. We further show that per-query cost prediction is fundamentally difficult: repeated runs of the same query yield thinking token variation up to 9.7x, establishing an irreducible noise floor for any predictor. Thus, we propose cost distribution prediction as an open challenge. Our findings demonstrate that listed API pricing is an unreliable proxy for actual cost, calling for cost-aware model selection and transparent per-request cost monitoring.

Dynamic LLM routers promise to cut inference costs by sending each query to the cheapest model that can answer it correctly. We analyze six commercial routers across 14 settings on a diverse benchmark spanning eight task categories, finding that none of them outperforms a router that randomly selects between two well-chosen models at matched cost. Some underperform by more than 10 percentage points. We trace this gap to four patterns prevalent across routers: difficulty blindness, length reversal, semantic matching, and roster suboptimality. We show that the first three are what the standard objective rewards: cost-accuracy Pareto efficiency on realized costs favors escalating moderately hard queries over the hardest ones, shorter queries over longer ones, and routing by a query's source over its difficulty. We also argue that the two assumptions that would justify large rosters, model granularity and model specialization, do not hold empirically. We propose an alternative evaluation methodology that does not reward these patterns, and as a proof of concept, we design a simple two-model router that avoids all four. Nevertheless, its gain over random routing is limited, because a well-chosen roster leaves little to route.

Routing benchmarks often judge a router against a hindsight oracle that counts a query as solved whenever any model's recorded answer is correct, but under stochastic decoding that oracle mixes two different objects: the clairvoyant ceiling of committing to one model per query and the premium of calling several models and keeping any correct answer. A companion study measured this premium at 3.5 percentage points (pp) for 11 small open models on MATH-500; how large it is once current reasoning models enter the pool was unknown. We measure it with a fixed-protocol, pre-committed generation on a single NVIDIA B300: all 500 MATH-500 problems, 30 seeded samples per problem, Llama-3.3-70B-Instruct and NVIDIA's Nemotron Nano 9B v2 and Nemotron Super 49B v1.5 (a NAS derivative of the same 70B) under vendor sampling, a 65,536-token cap, a pinned Math-Verify scorer, and an append-only attempt ledger, for 45,000 scored responses. The two reasoning models both reach 96.1% accuracy against 76.2% for the 70B baseline. The empirical product-coupling premium is 0.215-0.588 pp across all two- and three-model pools; the empirical plug-in selection headroom over always calling Nano is 1.27 pp; and resampling Nano three times gives practically the same empirical any-correct rate as calling all three models once. Only 31 responses reached the cap, 25 of them with exactly periodic tails, yet the capped responses account for up to 60.2% of a model's output-length-squared proxy, and long generations made a fourth planned model, Llama-3.1-8B, too slow to finish. The premium is far below the companion's 3.5 pp, but models, pool size and protocol changed together; an oracle gap should therefore be reported with the pool, the protocol, the samples per cell and the decoding budget that produced it.