Skip to content

Official Codebase of 3DCodeVerse

Python

pricing: one price table — each Price row carries its provenance, PROVENANCE goes

gaoypeng70bf338c610b
diff --git a/harness/codeverse3d/cli/cost_cmd.py b/harness/codeverse3d/cli/cost_cmd.py
index ffc9c039..70652500 100644
--- a/harness/codeverse3d/cli/cost_cmd.py
+++ b/harness/codeverse3d/cli/cost_cmd.py
@@ -168,7 +168,7 @@ def prices(
rows.append((f"{prov}:{model}", f"{p.input:g}", f"{p.cached:g}", f"{p.output:g}",
f"{p.image_usd:g}" if p.image_usd else "-", row.status, row.checked,
"-" if age is None else str(age), ",".join(problems) or "-",
- (row.provenance.note if row.provenance else "")[:52]))
+ p.note[:52]))
header = ("model", "in/M", "cached/M", "out/M", "$/img", "status", "checked", "age_d", "flags", "note")
widths = [max(len(str(r[i])) for r in [header, *rows]) for i in range(len(header))] if rows else \
[len(h) for h in header]
diff --git a/harness/codeverse3d/models/pricing.py b/harness/codeverse3d/models/pricing.py
index c39764e9..4d8d52c7 100644
--- a/harness/codeverse3d/models/pricing.py
+++ b/harness/codeverse3d/models/pricing.py
@@ -1,9 +1,9 @@
"""USD price table (per 1M tokens) + ``estimate_cost`` + price provenance.
-Every ``Usage`` is priced here (providers never return USD). Each row carries a
-:class:`Provenance` entry — where the number came from and when it was last
-checked — so a cost report can say *which* price row produced a dollar and
-whether it was verified or merely inferred (``3dcode cost prices``).
+Every ``Usage`` is priced here (providers never return USD). Each row carries its
+provenance — where the number came from and when it was last checked — so a cost
+report can say *which* price row produced a dollar and whether it was verified or
+merely inferred (``3dcode cost prices``).
Only the standard (short-context) tier is modelled, **except** the documented
>200k-prompt tiers of the Gemini pro models, which ``estimate_cost`` applies
@@ -27,7 +27,7 @@ log = logging.getLogger(__name__)
#: date the tables below were last reconciled against the providers' pricing pages
CHECKED = "2026-08-23"
-#: source ids used by :data:`PROVENANCE`
+#: ``Price.source`` ids; a row with no source was read off its provider's own page
SOURCES: dict[str, str] = {
"gemini": "https://ai.google.dev/gemini-api/docs/pricing",
"anthropic": "https://platform.claude.com/docs/en/about-claude/pricing",
@@ -49,10 +49,17 @@ class LongContext:
@dataclass(frozen=True)
class Price:
- """USD per 1M tokens. ``cached`` = cache-read input; ``cache_write`` = cache
- creation (Anthropic explicit caching, 5-minute TTL); ``thoughts`` are billed
- at the output rate. ``image_usd`` = per generated 1024² image for
- image-output models (their image tokens are billed per image, not per token).
+ """USD per 1M tokens, and where the number came from. ``cached`` = cache-read input;
+ ``cache_write`` = cache creation (Anthropic explicit caching, 5-minute TTL);
+ ``thoughts`` are billed at the output rate. ``image_usd`` = per generated 1024² image
+ for image-output models (their image tokens are billed per image, not per token).
+
+ Provenance: ``source`` is a :data:`SOURCES` key ("" = the provider's own pricing page),
+ ``checked`` the day it was last read there, and ``status``:
+
+ * ``verified`` — read off the provider's public pricing page on ``checked``
+ * ``inferred`` — no published row; copied from the nearest sibling
+ * ``unverified`` — the model is not listed on the page (retired / alias / CLI-only)
"""
input: float
@@ -62,91 +69,101 @@ class Price:
approximate: bool = False
image_usd: float = 0.0
long_context: LongContext | None = None
-
-
-@dataclass(frozen=True)
-class Provenance:
- """Where one price row came from. ``status``:
-
- * ``verified`` — read off the provider's public pricing page on ``checked``
- * ``inferred`` — no published row; copied from the nearest sibling
- * ``unverified`` — the model is not listed on the page (retired / alias / CLI-only)
- """
-
- source: str
+ source: str = ""
checked: str = CHECKED
status: str = "verified"
note: str = ""
- @property
- def url(self) -> str:
- return SOURCES.get(self.source, self.source)
-
-
-_V = Provenance # verified rows read straight off a pricing page
-
# fmt: off
PRICES: dict[tuple[str, str], Price] = {
# ---------------------------------------------------------------- gemini
# 3.6 / 3.7 / 3.8 flash: introductory rates valid through 2026-12-31; they double
# on 2027-01-01 ($1.50 / $7.50 / $0.15). Update then. No >200k surcharge.
- ("gemini", "gemini-3.8-flash"): Price(0.75, 3.75, 0.075),
- ("gemini", "gemini-3.7-flash"): Price(0.75, 3.75, 0.075),
- ("gemini", "gemini-3.6-flash"): Price(0.75, 3.75, 0.075),
+ ("gemini", "gemini-3.8-flash"): Price(0.75, 3.75, 0.075, checked="2026-09-21",
+ note="intro rate through 2026-12-31, then 1.50/7.50/0.15"),
+ ("gemini", "gemini-3.7-flash"): Price(0.75, 3.75, 0.075,
+ note="intro rate through 2026-12-31, then 1.50/7.50/0.15"),
+ ("gemini", "gemini-3.6-flash"): Price(0.75, 3.75, 0.075, note="intro rate through 2026-12-31"),
("gemini", "gemini-3.5-flash"): Price(1.50, 9.00, 0.15),
("gemini", "gemini-3.1-pro-preview"): Price(2.00, 12.00, 0.20,
- long_context=LongContext(200_000, 4.00, 18.00, 0.40)),
- ("gemini", "gemini-3.1-flash-lite"): Price(0.25, 1.50, 0.025),
+ long_context=LongContext(200_000, 4.00, 18.00, 0.40),
+ note=">200k prompts: 4.00/18.00/0.40 (modelled)"),
+ ("gemini", "gemini-3.1-flash-lite"): Price(0.25, 1.50, 0.025, note="audio input is 2x; text/image/video modelled"),
# image-output model: text/thinking output $3/M, IMAGE output $60/M ≈ $0.067
# per 1K image (1290 tokens); 0.5K $0.045, 2K $0.134, 4K $0.151. No cache row.
- ("gemini", "gemini-3.1-flash-image"): Price(0.50, 3.00, 0.05, image_usd=0.067),
+ ("gemini", "gemini-3.1-flash-image"): Price(0.50, 3.00, 0.05, image_usd=0.067,
+ note="image output $60/M ≈ $0.067 per 1K image; 0.5K $0.045, 4K $0.151"),
("gemini", "gemini-3-pro-preview"): Price(2.00, 12.00, 0.20,
- long_context=LongContext(200_000, 4.00, 18.00, 0.40)),
+ long_context=LongContext(200_000, 4.00, 18.00, 0.40),
+ note=">200k prompts: 4.00/18.00 (modelled)"),
("gemini", "gemini-3-flash-preview"): Price(0.50, 3.00, 0.05),
("gemini", "gemini-2.5-pro"): Price(1.25, 10.00, 0.125,
- long_context=LongContext(200_000, 2.50, 15.00, 0.25)),
- ("gemini", "gemini-2.5-flash"): Price(0.30, 2.50, 0.03),
- ("gemini", "gemini-2.5-flash-lite"): Price(0.10, 0.40, 0.01),
+ long_context=LongContext(200_000, 2.50, 15.00, 0.25),
+ note=">200k prompts: 2.50/15.00/0.25 (modelled)"),
+ ("gemini", "gemini-2.5-flash"): Price(0.30, 2.50, 0.03, note="cached was 0.075 here until 2026-08-23; published 0.03"),
+ ("gemini", "gemini-2.5-flash-lite"): Price(0.10, 0.40, 0.01, note="cached was 0.025 here until 2026-08-23; published 0.01"),
# image tokens are billed per image: 1024² = 1290 tokens = $0.039 ⇒ $30.23/M
- ("gemini", "gemini-2.5-flash-image"): Price(0.30, 30.00, 0.03, image_usd=0.039),
+ ("gemini", "gemini-2.5-flash-image"): Price(0.30, 30.00, 0.03, image_usd=0.039, status="inferred",
+ note="output/M derived from $0.039 per 1290-token 1024² image"),
# ------------------------------------------------------------- anthropic
- ("anthropic", "claude-fable-5"): Price(10.00, 50.00, 1.00, cache_write=12.50),
- ("anthropic", "claude-mythos-5"): Price(10.00, 50.00, 1.00, cache_write=12.50),
- ("anthropic", "claude-opus-5"): Price(5.00, 25.00, 0.50, cache_write=6.25),
+ ("anthropic", "claude-fable-5"): Price(10.00, 50.00, 1.00, cache_write=12.50,
+ note="1h cache write 20.00 (not modelled)"),
+ ("anthropic", "claude-mythos-5"): Price(10.00, 50.00, 1.00, cache_write=12.50, note="limited availability"),
+ ("anthropic", "claude-opus-5"): Price(5.00, 25.00, 0.50, cache_write=6.25,
+ note="fast mode is 10.00/50.00 (not modelled)"),
# the 1M-context variant of opus-5, and the id the DEFAULT claude-code arm serves.
# "[1m]" is not a version suffix, so it cannot prefix-match claude-opus-5 and priced
# at $0.00 until 2026-08-24 (a recorded run billed $1.218 under this exact key).
- ("anthropic", "claude-opus-5[1m]"): Price(5.00, 25.00, 0.50, cache_write=6.25),
- ("anthropic", "claude-opus-4-8"): Price(5.00, 25.00, 0.50, cache_write=6.25),
+ ("anthropic", "claude-opus-5[1m]"): Price(5.00, 25.00, 0.50, cache_write=6.25, source="provider-cost",
+ checked="2026-08-24", status="inferred",
+ note="1M-context opus-5, served by the default claude-code arm. A probe "
+ "billed costUSD 0.034620 for in=2, out=4, 1h-cache-write=3451 — "
+ "exactly 2x5.00 + 4x25.00 + 3451x10.00 per 1M, i.e. the standard "
+ "opus-5 rates with the 1h cache write at 2x input (not modelled here, "
+ "as for every other Anthropic row)"),
+ ("anthropic", "claude-opus-4-8"): Price(5.00, 25.00, 0.50, cache_write=6.25,
+ note="fast mode is 10.00/50.00 (not modelled)"),
("anthropic", "claude-opus-4-7"): Price(5.00, 25.00, 0.50, cache_write=6.25),
("anthropic", "claude-opus-4-6"): Price(5.00, 25.00, 0.50, cache_write=6.25),
("anthropic", "claude-opus-4-5"): Price(5.00, 25.00, 0.50, cache_write=6.25),
- ("anthropic", "claude-opus-4-1"): Price(15.00, 75.00, 1.50, cache_write=18.75),
- ("anthropic", "claude-opus-4"): Price(15.00, 75.00, 1.50, cache_write=18.75),
+ ("anthropic", "claude-opus-4-1"): Price(15.00, 75.00, 1.50, cache_write=18.75,
+ note="retired except on Bedrock / Google Cloud"),
+ ("anthropic", "claude-opus-4"): Price(15.00, 75.00, 1.50, cache_write=18.75, note="retired except on Google Cloud"),
# sonnet-5: the $2/$10 launch rate became the standard rate (the 2026-09-01
# increase to $3/$15 was cancelled) — was priced 50% too high here until 2026-08-23.
- ("anthropic", "claude-sonnet-5"): Price(2.00, 10.00, 0.20, cache_write=2.50),
+ ("anthropic", "claude-sonnet-5"): Price(2.00, 10.00, 0.20, cache_write=2.50,
+ note="launch rate 2.00/10.00 is now standard; was 3.00/15.00 here"),
("anthropic", "claude-sonnet-4-6"): Price(3.00, 15.00, 0.30, cache_write=3.75),
("anthropic", "claude-sonnet-4-5"): Price(3.00, 15.00, 0.30, cache_write=3.75),
- ("anthropic", "claude-sonnet-4"): Price(3.00, 15.00, 0.30, cache_write=3.75),
+ ("anthropic", "claude-sonnet-4"): Price(3.00, 15.00, 0.30, cache_write=3.75,
+ note="retired except on Bedrock / Google Cloud"),
("anthropic", "claude-haiku-4-5"): Price(1.00, 5.00, 0.10, cache_write=1.25),
- ("anthropic", "claude-haiku-4"): Price(1.00, 5.00, 0.10, cache_write=1.25, approximate=True),
- ("anthropic", "claude-haiku-3-5"): Price(0.80, 4.00, 0.08, cache_write=1.00),
+ ("anthropic", "claude-haiku-4"): Price(1.00, 5.00, 0.10, cache_write=1.25, approximate=True, source="inferred",
+ status="unverified",
+ note="not listed on the pricing page; assumed = haiku-4.5"),
+ ("anthropic", "claude-haiku-3-5"): Price(0.80, 4.00, 0.08, cache_write=1.00,
+ note="retired except on Bedrock / Google Cloud"),
# ---------------------------------------------------------------- openai
# gpt-5.6-sol is CHEAPER than gpt-5.6 (was priced as its equal until 2026-08-23).
- ("openai", "gpt-5.6-sol"): Price(4.00, 20.00, 0.40),
- ("openai", "gpt-5.6-terra"): Price(2.00, 12.00, 0.20),
- ("openai", "gpt-5.6-luna"): Price(0.20, 1.20, 0.02),
+ ("openai", "gpt-5.6-sol"): Price(4.00, 20.00, 0.40,
+ note="4.00/20.00/0.40; was 5.00/30.00/0.50 here until 2026-08-23"),
+ ("openai", "gpt-5.6-terra"): Price(2.00, 12.00, 0.20, checked="2026-08-24",
+ note="mid tier; fast mode is 4.00/24.00 (not modelled)"),
+ ("openai", "gpt-5.6-luna"): Price(0.20, 1.20, 0.02, checked="2026-08-24",
+ note="small tier, cut 80% on 2026-07-30; fast mode is 0.40/2.40 (not modelled)"),
("openai", "gpt-5.6"): Price(5.00, 30.00, 0.50),
("openai", "gpt-5.5"): Price(5.00, 30.00, 0.50),
("openai", "gpt-5.4"): Price(2.50, 15.00, 0.25),
("openai", "gpt-5.3-codex"): Price(1.75, 14.00, 0.175),
- ("openai", "gpt-5.2-codex"): Price(1.75, 14.00, 0.175, approximate=True),
+ ("openai", "gpt-5.2-codex"): Price(1.75, 14.00, 0.175, approximate=True, source="inferred",
+ status="inferred", note="not listed; assumed = gpt-5.2"),
("openai", "gpt-5.2"): Price(1.75, 14.00, 0.175),
- ("openai", "gpt-5.1-codex"): Price(1.25, 10.00, 0.125, approximate=True),
+ ("openai", "gpt-5.1-codex"): Price(1.25, 10.00, 0.125, approximate=True, source="inferred",
+ status="inferred", note="not listed; assumed = gpt-5.1"),
("openai", "gpt-5.1"): Price(1.25, 10.00, 0.125),
- ("openai", "gpt-5-codex"): Price(1.25, 10.00, 0.125, approximate=True),
+ ("openai", "gpt-5-codex"): Price(1.25, 10.00, 0.125, approximate=True, source="inferred",
+ status="inferred", note="not listed; assumed = gpt-5"),
("openai", "gpt-5-mini"): Price(0.25, 2.00, 0.025),
("openai", "gpt-5-nano"): Price(0.05, 0.40, 0.005),
("openai", "gpt-5"): Price(1.25, 10.00, 0.125),
@@ -157,81 +174,11 @@ PRICES: dict[tuple[str, str], Price] = {
("openai", "gpt-4o-mini"): Price(0.15, 0.60, 0.075),
("openai", "o3"): Price(2.00, 8.00, 0.50),
("openai", "o3-mini"): Price(1.10, 4.40, 0.55),
- ("openai", "o3-pro"): Price(20.00, 80.00),
+ ("openai", "o3-pro"): Price(20.00, 80.00, note="no cached-input rate published"),
("openai", "o4-mini"): Price(1.10, 4.40, 0.275),
("openai", "o1"): Price(15.00, 60.00, 7.50),
- ("openai", "o1-mini"): Price(1.10, 4.40, 0.55, approximate=True),
-}
-
-#: per price row: where the number came from and when it was last checked.
-#: Every key of :data:`PRICES` must appear here (``tests/cost/test_price_hygiene.py``).
-PROVENANCE: dict[tuple[str, str], Provenance] = {
- ("gemini", "gemini-3.8-flash"): _V("gemini", checked="2026-09-21",
- note="intro rate through 2026-12-31, then 1.50/7.50/0.15"),
- ("gemini", "gemini-3.7-flash"): _V("gemini", note="intro rate through 2026-12-31, then 1.50/7.50/0.15"),
- ("gemini", "gemini-3.6-flash"): _V("gemini", note="intro rate through 2026-12-31"),
- ("gemini", "gemini-3.5-flash"): _V("gemini"),
- ("gemini", "gemini-3.1-pro-preview"): _V("gemini", note=">200k prompts: 4.00/18.00/0.40 (modelled)"),
- ("gemini", "gemini-3.1-flash-lite"): _V("gemini", note="audio input is 2x; text/image/video modelled"),
- ("gemini", "gemini-3.1-flash-image"): _V("gemini", note="image output $60/M ≈ $0.067 per 1K image; 0.5K $0.045, 4K $0.151"),
- ("gemini", "gemini-3-pro-preview"): _V("gemini", note=">200k prompts: 4.00/18.00 (modelled)"),
- ("gemini", "gemini-3-flash-preview"): _V("gemini"),
- ("gemini", "gemini-2.5-pro"): _V("gemini", note=">200k prompts: 2.50/15.00/0.25 (modelled)"),
- ("gemini", "gemini-2.5-flash"): _V("gemini", note="cached was 0.075 here until 2026-08-23; published 0.03"),
- ("gemini", "gemini-2.5-flash-lite"): _V("gemini", note="cached was 0.025 here until 2026-08-23; published 0.01"),
- ("gemini", "gemini-2.5-flash-image"): _V("gemini", status="inferred",
- note="output/M derived from $0.039 per 1290-token 1024² image"),
- ("anthropic", "claude-fable-5"): _V("anthropic", note="1h cache write 20.00 (not modelled)"),
- ("anthropic", "claude-mythos-5"): _V("anthropic", note="limited availability"),
- ("anthropic", "claude-opus-5"): _V("anthropic", note="fast mode is 10.00/50.00 (not modelled)"),
- ("anthropic", "claude-opus-5[1m]"): _V("provider-cost", checked="2026-08-24", status="inferred",
- note="1M-context opus-5, served by the default claude-code arm. A probe "
- "billed costUSD 0.034620 for in=2, out=4, 1h-cache-write=3451 — "
- "exactly 2x5.00 + 4x25.00 + 3451x10.00 per 1M, i.e. the standard "
- "opus-5 rates with the 1h cache write at 2x input (not modelled here, "
- "as for every other Anthropic row)"),
- ("anthropic", "claude-opus-4-8"): _V("anthropic", note="fast mode is 10.00/50.00 (not modelled)"),
- ("anthropic", "claude-opus-4-7"): _V("anthropic"),
- ("anthropic", "claude-opus-4-6"): _V("anthropic"),
- ("anthropic", "claude-opus-4-5"): _V("anthropic"),
- ("anthropic", "claude-opus-4-1"): _V("anthropic", note="retired except on Bedrock / Google Cloud"),
- ("anthropic", "claude-opus-4"): _V("anthropic", note="retired except on Google Cloud"),
- ("anthropic", "claude-sonnet-5"): _V("anthropic", note="launch rate 2.00/10.00 is now standard; was 3.00/15.00 here"),
- ("anthropic", "claude-sonnet-4-6"): _V("anthropic"),
- ("anthropic", "claude-sonnet-4-5"): _V("anthropic"),
- ("anthropic", "claude-sonnet-4"): _V("anthropic", note="retired except on Bedrock / Google Cloud"),
- ("anthropic", "claude-haiku-4-5"): _V("anthropic"),
- ("anthropic", "claude-haiku-4"): _V("inferred", status="unverified",
- note="not listed on the pricing page; assumed = haiku-4.5"),
- ("anthropic", "claude-haiku-3-5"): _V("anthropic", note="retired except on Bedrock / Google Cloud"),
- ("openai", "gpt-5.6-sol"): _V("openai", note="4.00/20.00/0.40; was 5.00/30.00/0.50 here until 2026-08-23"),
- ("openai", "gpt-5.6-terra"): _V("openai", checked="2026-08-24",
- note="mid tier; fast mode is 4.00/24.00 (not modelled)"),
- ("openai", "gpt-5.6-luna"): _V("openai", checked="2026-08-24",
- note="small tier, cut 80% on 2026-07-30; fast mode is 0.40/2.40 (not modelled)"),
- ("openai", "gpt-5.6"): _V("openai"),
- ("openai", "gpt-5.5"): _V("openai"),
- ("openai", "gpt-5.4"): _V("openai"),
- ("openai", "gpt-5.3-codex"): _V("openai"),
- ("openai", "gpt-5.2-codex"): _V("inferred", status="inferred", note="not listed; assumed = gpt-5.2"),
- ("openai", "gpt-5.2"): _V("openai"),
- ("openai", "gpt-5.1-codex"): _V("inferred", status="inferred", note="not listed; assumed = gpt-5.1"),
- ("openai", "gpt-5.1"): _V("openai"),
- ("openai", "gpt-5-codex"): _V("inferred", status="inferred", note="not listed; assumed = gpt-5"),
- ("openai", "gpt-5-mini"): _V("openai"),
- ("openai", "gpt-5-nano"): _V("openai"),
- ("openai", "gpt-5"): _V("openai"),
- ("openai", "gpt-4.1"): _V("openai"),
- ("openai", "gpt-4.1-mini"): _V("openai"),
- ("openai", "gpt-4.1-nano"): _V("openai"),
- ("openai", "gpt-4o"): _V("openai"),
- ("openai", "gpt-4o-mini"): _V("openai"),
- ("openai", "o3"): _V("openai"),
- ("openai", "o3-mini"): _V("openai"),
- ("openai", "o3-pro"): _V("openai", note="no cached-input rate published"),
- ("openai", "o4-mini"): _V("openai"),
- ("openai", "o1"): _V("openai"),
- ("openai", "o1-mini"): _V("inferred", status="unverified", note="not listed; kept at the o3-mini rate"),
+ ("openai", "o1-mini"): Price(1.10, 4.40, 0.55, approximate=True, source="inferred",
+ status="unverified", note="not listed; kept at the o3-mini rate"),
}
# fmt: on
@@ -297,29 +244,20 @@ class PriceRow:
key: str # the PRICES key that matched ("" when unknown)
match: str # exact | prefix | unknown
price: Price | None
- provenance: Provenance | None
@property
def approximate(self) -> bool:
"""True when the dollar cannot be trusted to the cent: no row, an
inferred/unverified row, or a row flagged ``approximate``."""
- if self.price is None or self.provenance is None:
- return True
- return self.price.approximate or self.provenance.status != "verified"
-
- @property
- def source(self) -> str:
- return self.provenance.url if self.provenance else "unknown"
+ return self.price is None or self.price.approximate or self.price.status != "verified"
@property
def checked(self) -> str:
- return self.provenance.checked if self.provenance else ""
+ return self.price.checked if self.price else ""
@property
def status(self) -> str:
- if self.price is None:
- return "unknown"
- return self.provenance.status if self.provenance else "unverified"
+ return self.price.status if self.price else "unknown"
def price_provenance(provider: str, model: str) -> PriceRow:
@@ -328,10 +266,10 @@ def price_provenance(provider: str, model: str) -> PriceRow:
prov = (provider or "").strip().lower()
found = _match(prov, model)
if not found:
- return PriceRow(prov, _normalise(model), "", "unknown", None, None)
+ return PriceRow(prov, _normalise(model), "", "unknown", None)
key, price = found
match = "exact" if key[1] == _normalise(model) else "prefix"
- return PriceRow(prov, _normalise(model), key[1], match, price, PROVENANCE.get(key))
+ return PriceRow(prov, _normalise(model), key[1], match, price)
def unit_prices(provider: str, model: str, *, prompt_tokens: int = 0) -> tuple[float, float, float]:
diff --git a/harness/docs/COST.md b/harness/docs/COST.md
index 16b15a45..619a6308 100644
--- a/harness/docs/COST.md
+++ b/harness/docs/COST.md
@@ -280,9 +280,10 @@ recorded before it.
## 7. Price hygiene (checked 2026-08-23)
-`models/pricing.py` now carries a `PROVENANCE` row per price — source URL,
-`checked` date and a status of `verified` / `inferred` / `unverified` — exposed
-as `price_provenance(provider, model)` and printed by `3dcode cost prices`. Every
+Every row of `models/pricing.PRICES` carries its provenance — source, `checked` date and a
+status of `verified` / `inferred` / `unverified` (one table since 2026-09-22; the parallel
+`PROVENANCE` dict is gone) — exposed as `price_provenance(provider, model)` and printed by
+`3dcode cost prices`. Every
row was reconciled against the providers' live pricing pages. **Corrections
made:**
diff --git a/harness/tests/cost/test_price_hygiene.py b/harness/tests/cost/test_price_hygiene.py
index aefb2226..0cafc7ce 100644
--- a/harness/tests/cost/test_price_hygiene.py
+++ b/harness/tests/cost/test_price_hygiene.py
@@ -8,7 +8,7 @@ from codeverse3d.contracts.common import Usage
from codeverse3d.models.pricing import (
IMAGE_USD_BY_SIZE,
PRICES,
- PROVENANCE,
+ SOURCES,
estimate_cost,
per_image_usd,
price_provenance,
@@ -38,13 +38,10 @@ SUPPORTED_MODEL_IDS = (
def test_every_price_row_has_provenance():
- assert set(PRICES) == set(PROVENANCE), (
- f"missing provenance: {sorted(set(PRICES) - set(PROVENANCE))}; "
- f"stale provenance: {sorted(set(PROVENANCE) - set(PRICES))}")
- for key, prov in PROVENANCE.items():
- assert prov.checked and prov.checked[:2] == "20", key
- assert prov.status in ("verified", "inferred", "unverified"), key
- assert prov.url, key
+ for (provider, model), price in PRICES.items():
+ assert price.checked and price.checked[:2] == "20", model
+ assert price.status in ("verified", "inferred", "unverified"), model
+ assert (price.source or provider) in SOURCES, model
def test_provenance_flags_the_rows_we_cannot_stand_behind():
@@ -90,7 +87,7 @@ def test_the_bracketed_context_variant_prices_on_its_own_row():
"""A bracketed context variant requires its own exact price row."""
row = price_provenance("anthropic", "claude-opus-5[1m]")
assert row.match == "exact" and row.price is not None
- assert row.provenance.status == "inferred" and row.provenance.checked >= "2026-08-24"
+ assert row.status == "inferred" and row.checked >= "2026-08-24"
u = Usage(input_tokens=2, output_tokens=46_921)
assert estimate_cost("anthropic", "claude-opus-5[1m]", u) == pytest.approx(
estimate_cost("anthropic", "claude-opus-5", u)), "same rates as the standard-context row"
@@ -101,7 +98,7 @@ def test_the_bracketed_context_variant_prices_on_its_own_row():
# --------------------------------------------------------------- staleness maintenance
def test_every_price_row_was_checked_within_the_maintenance_window():
"""A price nobody re-checked for 90 days is not evidence. When this fails,
- re-read the providers' pricing pages, update PRICES/PROVENANCE and bump
+ re-read the providers' pricing pages, update PRICES and bump
``CHECKED`` — do not raise the threshold."""
from datetime import date, datetime