GPT 5.6 Terra

Name: GPT 5.6 Terra — independent benchmark scores
Creator: Epoch AI
License: https://creativecommons.org/licenses/by/4.0/

OpenAIChat1.1M context· released 2026-07-09Floating alias

gpt-5.6-terra

Estimate cost Compare

Prices as of 2026-07-14 · rev 179affb· unknown is shown as unknown, never $0.00

Strongest atReasoning · Mathematicstop-quartile among comparable models scored on these domains

Value position95th pctile GPQA3 cheaper models match or beat it on GPQA — see the value chart

Data confidence4/4 exactbenchmarks measured on this exact model · prices 2026-07-14

Cost / request$0.011K in · 500 out

Input / M$2.50

Output / M$15.00

Cache read / M$0.25

Context1.1M1,050,000 tokens

Max output128K128,000 tokens

Capability fingerprint

independent benchmarks, grouped by domain · higher is better · not our measurement

Reasoning

GPQA Diamondgraduate-level science reasoning

91%95th pctile

exact modeloptimized runEpoch evaluations · GPT-5.6 Terra

SimpleBencheveryday reasoning (unsaturated)

39%62nd pctile

exact modelSimpleBench Leaderboard · GPT-5.6 Terra

Mathematics

AIME (OTIS mock)olympiad math

100%98th pctile

exact modeloptimized runEpoch evaluations · GPT-5.6 Terra

Factuality

SimpleQAfactual accuracy (higher = fewer hallucinations)

43%57th pctile

exact modelEpoch evaluations · GPT-5.6 Terra

peer median · percentile ranks this model among distinct comparable models (deduped — one entry per model, not per provider) scored on that benchmark · exact model / family proxy per row.

Source: Epoch AI — “AI Benchmarking Hub” (epoch.ai), CC-BY (Epoch AI). Some rows carry an upstream leaderboard (shown per row); Aider & Terminal-Bench are additionally Apache-2.0. Each benchmark measures a different thing on a different scale — rows are not comparable across domains, and a score is not a guarantee of quality on your task.

Capabilities Index

Epoch AI's cross-model capability re-fit · one comparable axis · not our measurement

158.2ECI

97th percentile of 150 modelsexact model

The dot is GPT 5.6 Terra; the whisker is its 95% CI; the firm tick is the peer median; faint ticks are every other scored model. Unlike a raw benchmark, the ECI axis is comparable across models — but it's a relative re-fit scale, so a model's exact position can shift when Epoch recomputes it.

Source: Epoch AI — Capabilities Index & Notable-Models data (epoch.ai), CC-BY (Epoch AI). The ECI is Epoch's index, not TokenTriage's; per-fact confidence labels are Epoch's own.

Cost vs quality

where GPT 5.6 Terra sits among GPQA-scored models

Each dot is a model with a GPQA Diamond score, priced at a 1K-in / 500-out request; GPT 5.6 Terra is highlighted. The shaded corner holds models that are both cheaper and higher-scoring; the dashed staircase is the value frontier. Hover any dot for its model.

3 cheaper models match or beat GPT 5.6 Terra on GPQA— see the full table below, switchable by benchmark.

Also worth considering

cheaper models that score about as well — pick the benchmark that matters to you

8 cheaper models score within 3 points of GPT 5.6 Terra on GPQA Diamond.

Model	GPQA Diamond	Cost / req	Savings
CLGlm 5.2	89%-2	$0.0036	−64%
AOGPT 5.6 Luna	89%-2	$0.004	−60%
XAGrok 4.5	91%+0	$0.005	−50%
XAGrok 4.5 Latest	91%+0	$0.005	−50%
GGGemini 3.5 Flash	90%-1	$0.006	−40%
GGGemini 3 Pro Preview	90%-1	$0.008	−20%
GGGemini 3.1 Pro Preview	92%+1	$0.008	−20%
REGemini 3 Pro	90%-1	$0.008	−20%

Deduped to one entry per model (cheapest offering), priced at the reference 1K-in / 500-out request. Filtered only on the benchmark score — a model with an unknown capability is never silently excluded. A lower price is only cheaper if quality holds on your task — that's what routing proves.

Pricing

Input · fresh prompt tokens	$2.50 / M✓ 3 sources agree
Output · generated tokens	$15.00 / M✓ 3 sources agree
Cache read · cached-input hit	$0.25 / M
Cache write · 5-minute TTL	$3.13 / M
Reasoning · thinking tokens	unknown
Batch input · async 50% tier	$1.25 / M
Batch output · async 50% tier	$7.50 / M

Context-length tiers

Base (≤ 272K)	$2.50 / $15.00 per M
Above 272K	$5.00 / $22.50 per M

When total input exceeds the threshold, the tier rate applies to the whole request.

Cross the 272K line by a single token and the whole request reprices — every token, not just those above the line. Assumes 500 output tokens (the cliff's position depends on input alone). Snapshot rates; providers do change tier structure.

Reasoning tokens: no separate rate is published. Providers typically bill thinking tokens at the output rate ($15.00 / M) — which is what the estimator assumes, so an extended-thinking workload isn't silently undercounted.

Model info

Provider: OpenAI
Catalog key: openai/gpt-5.6-terra
Identifier: Floating alias (may re-point over time)
Type: Chat
Max input: 1,050,000 tokens
Max output: 128,000 tokens
Released: 2026-07-09
Knowledge cutoff: 2026-02-16
Weights: Proprietary
Price snapshot: 2026-07-14 · rev 179affb

Pricing from TokenTriage's resolved price artifact (TokenTriage resolved price artifact (internal/pricing/data/prices.json)); capability + model metadata is third-party (LiteLLM model_prices_and_context_window.json, models.dev). Unknown means unstated, not absent.

Prompt-cache break-even

reuse the same prefix ≥ 2 times and caching wins

Writing a prefix to cache costs a premium, but each subsequent hit reads at $0.25/M instead of $2.50/M. For this model the crossover is 2 reuses — past that, the same prompt keeps getting cheaper. TokenTriage's semantic cache is built to cross that line for you.

Effective cost / 1,000 words

normalised comparison metric · your headline price is per token, but tokenizers pack words differently

gpt-5.6-terra uses the OpenAI o200k (GPT-4o / GPT-4.1 / o-series) tokenizer. The figures below multiply its per-token price by how many tokens that tokenizer produces for representative text — so two models at the samelist price aren't the same price per word. This is a comparison lens, not a bill:the estimator remains the authoritative cost for your own traffic.

⚠ The OpenAI o200k (GPT-4o / GPT-4.1 / o-series) tokenizer is inferred from OpenAI's current base encoding for this model — not independently confirmed. If it ships a different tokenizer, its per-word figures would shift.

$0.0029per 1K words · input

$0.017per 1K words · output

1,150tokens / 1K words

Among 75 models with a public tokenizer, GPT 5.6 Terra ranks 44th by list price per token and 44th by cost per 1,000 words — normalising doesn't change its position here.

Tokens/word measured with gpt-tokenizer (MIT) over a fixed self-authored corpus (as of 2026-07-29); only derived counts are stored, no vocabulary is shipped. Word count: whitespace-delimited tokens; approximate for non-spaced scripts (CJK) in the multilingual sample.Models whose tokenizer isn't publicly reproducible (e.g. Claude, Gemini) are not comparable here and are excluded — never assigned a guessed count.

Capabilities

9 documented as supported

Input & output

✓Vision

✓PDF input

?Audio input

?Audio output

API behaviour

✓Streaming

✓Structured output

✓Prompt caching

✓Reasoning

Agents & tools

✓Function calling

✓Parallel tool calls

✓Web search

?Computer use

✓ supported ·✕ not supported ·? not documented ·⚠ sources disagree. A models.dev tag means the LiteLLM catalog was silent and an independent source supplied the value (lower confidence); ? is never silently turned into a confident ✕.

Estimate a request

disjoint token buckets · input = fresh (uncached) tokens

Preset

Input tokens

Output tokens

Cached input

Cache write

Reasoning

Requests / month

Cost / request—

Projected / month—

How output length drives the bill

at a 1K-token prompt · cost scales linearly with output

Output tokens overtake input spend at ~167 tokens for this model — past that, generation is your bill, whatever the headline input price says. Computed by the same cost engine as the estimate above.

Compare GPT 5.6 Terra with other models →

Route to GPT 5.6 Terra

OpenAI-compatible · one base URL

curl https://YOUR-TOKENTRIAGE-HOST/v1/chat/completions \
  -H "Authorization: Bearer $TOKENTRIAGE_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-5.6-terra",
    "messages": [{"role": "user", "content": "Hello"}]
  }'from openai import OpenAI

client = OpenAI(
    base_url="https://YOUR-TOKENTRIAGE-HOST/v1",
    api_key="YOUR_TOKENTRIAGE_KEY",
)

resp = client.chat.completions.create(
    model="openai/gpt-5.6-terra",
    messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)

Point your existing OpenAI client at your TokenTriage host and pass the catalog key as the model. TokenTriage attributes the real cost per request and can route to a cheaper model once it's proven just as good on your traffic.

Data & API

machine-readable · rev 179affb

Pull this model's pricing, capabilities, benchmarks and cross-source provenance programmatically — every field carries the snapshot rev, and unknown is null, never 0. Free to reuse with attribution: benchmarks & Capabilities Index © Epoch AI (CC-BY (Epoch AI)); human-preference Elo © LMArena (CC-BY-4.0); capability & price cross-check via models.dev v2 (MIT) · Portkey (MIT) · TrueFoundry (MIT); open-weights facts via Hugging Face (per-repo license).

openai-gpt-5-6-terra.json openai-gpt-5-6-terra.csv

A price is a guess until it meets your traffic.

This page tells you what GPT 5.6 Terra charges per token. TokenTriage tells you what it costs on your real requests — and cuts the bill only after proving a cheaper model is just as good.

See the product Back to directory