From 0af33da5016446518a57d44336c6dfc66c87d0c2 Mon Sep 17 00:00:00 2001 From: pbhhdf Date: Wed, 16 Sep 2026 13:15:32 +0800 Subject: [PATCH 1/7] feat(gptzzz): add GPTZZZ provider with the GPT-5.6 and GPT-6 catalog GPTZZZ (https://gptzzz.ai) is an OpenAI-compatible relay. Adds the provider entry plus the four chat models it currently serves, all via base_model on the existing OpenAI lab entries. Sources: - Endpoint and catalog: https://gptzzz.ai/docs/ and GET /v1/models (2026-09-16) - Effort values verified live against POST /v1/chat/completions: none, low, medium, high, xhigh and max are accepted; minimal is rejected upstream. - Pricing is the published 0.2x group rate applied to OpenAI list pricing, cross-checked against the cost/actual_cost fields from GET /v1/usage. Co-Authored-By: Claude Opus 5 --- providers/gptzzz/logo.svg | 1 + providers/gptzzz/models/gpt-5.6-sol.toml | 20 ++++++++++++++++++++ providers/gptzzz/models/gpt-5.6-terra.toml | 20 ++++++++++++++++++++ providers/gptzzz/models/gpt-5.6.toml | 21 +++++++++++++++++++++ providers/gptzzz/models/gpt-6-astra.toml | 20 ++++++++++++++++++++ providers/gptzzz/provider.toml | 5 +++++ 6 files changed, 87 insertions(+) create mode 100644 providers/gptzzz/logo.svg create mode 100644 providers/gptzzz/models/gpt-5.6-sol.toml create mode 100644 providers/gptzzz/models/gpt-5.6-terra.toml create mode 100644 providers/gptzzz/models/gpt-5.6.toml create mode 100644 providers/gptzzz/models/gpt-6-astra.toml create mode 100644 providers/gptzzz/provider.toml diff --git a/providers/gptzzz/logo.svg b/providers/gptzzz/logo.svg new file mode 100644 index 00000000000..f8be18d2183 --- /dev/null +++ b/providers/gptzzz/logo.svg @@ -0,0 +1 @@ +GPTZZZ diff --git a/providers/gptzzz/models/gpt-5.6-sol.toml b/providers/gptzzz/models/gpt-5.6-sol.toml new file mode 100644 index 00000000000..af47ede30a4 --- /dev/null +++ b/providers/gptzzz/models/gpt-5.6-sol.toml @@ -0,0 +1,20 @@ +# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live +# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max +# are accepted; minimal is rejected upstream. +# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked +# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +base_model = "openai/gpt-5.6-sol" +reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] + +[cost] +input = 0.8 +output = 4.00 +cache_read = 0.08 +cache_write = 1.00 + +[[cost.tiers]] +tier = { size = 272_000 } +input = 1.6 +output = 6.00 +cache_read = 0.16 +cache_write = 2.00 diff --git a/providers/gptzzz/models/gpt-5.6-terra.toml b/providers/gptzzz/models/gpt-5.6-terra.toml new file mode 100644 index 00000000000..79ee439c64e --- /dev/null +++ b/providers/gptzzz/models/gpt-5.6-terra.toml @@ -0,0 +1,20 @@ +# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live +# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max +# are accepted; minimal is rejected upstream. +# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked +# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +base_model = "openai/gpt-5.6-terra" +reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] + +[cost] +input = 0.4 +output = 2.4 +cache_read = 0.04 +cache_write = 0.5 + +[[cost.tiers]] +tier = { size = 272_000 } +input = 0.8 +output = 3.6 +cache_read = 0.08 +cache_write = 1.00 diff --git a/providers/gptzzz/models/gpt-5.6.toml b/providers/gptzzz/models/gpt-5.6.toml new file mode 100644 index 00000000000..10e2e58d1bf --- /dev/null +++ b/providers/gptzzz/models/gpt-5.6.toml @@ -0,0 +1,21 @@ +# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live +# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max +# are accepted; minimal is rejected upstream. +# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked +# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +base_model = "openai/gpt-5.6-sol" +name = "GPT-5.6" +reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] + +[cost] +input = 0.8 +output = 4.00 +cache_read = 0.08 +cache_write = 1.00 + +[[cost.tiers]] +tier = { size = 272_000 } +input = 1.6 +output = 6.00 +cache_read = 0.16 +cache_write = 2.00 diff --git a/providers/gptzzz/models/gpt-6-astra.toml b/providers/gptzzz/models/gpt-6-astra.toml new file mode 100644 index 00000000000..17c1dee9685 --- /dev/null +++ b/providers/gptzzz/models/gpt-6-astra.toml @@ -0,0 +1,20 @@ +# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live +# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max +# are accepted; minimal is rejected upstream. +# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked +# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +base_model = "openai/gpt-6-astra" +reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] + +[cost] +input = 2.00 +output = 10.00 +cache_read = 0.2 +cache_write = 2.5 + +[[cost.tiers]] +tier = { size = 272_000 } +input = 4.00 +output = 15.00 +cache_read = 0.4 +cache_write = 5.00 diff --git a/providers/gptzzz/provider.toml b/providers/gptzzz/provider.toml new file mode 100644 index 00000000000..c8f9cd3abee --- /dev/null +++ b/providers/gptzzz/provider.toml @@ -0,0 +1,5 @@ +name = "GPTZZZ" +env = ["GPTZZZ_API_KEY"] +npm = "@ai-sdk/openai-compatible" +api = "https://gptzzz.ai/v1" +doc = "https://gptzzz.ai/docs/" From 1df70787f96ccadc96c4a36dc3c70dc86c0482dd Mon Sep 17 00:00:00 2001 From: pbhhdf Date: Wed, 16 Sep 2026 14:38:23 +0800 Subject: [PATCH 2/7] fix(gptzzz): drop none effort on gpt-6-astra, verify effort per model Review flagged that the gpt-6-astra effort list was copied from GPT-5.6 rather than verified on that model. Re-tested each model separately against the live endpoint. gpt-6-astra rejects none upstream: Unsupported value: 'none' is not supported with the 'gpt-6-astra' model. Supported values are: 'low', 'medium', 'high', 'xhigh', and 'max'. This now matches providers/openai/models/gpt-6-astra.toml and the other same-surface peers. gpt-5.6, gpt-5.6-sol and gpt-5.6-terra do accept none (HTTP 200), so their lists are unchanged; their header comments now say the values were verified per model. Co-Authored-By: Claude Opus 5 --- providers/gptzzz/models/gpt-5.6-sol.toml | 6 +++--- providers/gptzzz/models/gpt-5.6-terra.toml | 6 +++--- providers/gptzzz/models/gpt-5.6.toml | 6 +++--- providers/gptzzz/models/gpt-6-astra.toml | 9 +++++---- 4 files changed, 14 insertions(+), 13 deletions(-) diff --git a/providers/gptzzz/models/gpt-5.6-sol.toml b/providers/gptzzz/models/gpt-5.6-sol.toml index af47ede30a4..ae55ff1c1a3 100644 --- a/providers/gptzzz/models/gpt-5.6-sol.toml +++ b/providers/gptzzz/models/gpt-5.6-sol.toml @@ -1,6 +1,6 @@ -# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live -# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max -# are accepted; minimal is rejected upstream. +# GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the +# live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh +# and max are all accepted on this model; minimal is rejected upstream. # Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked # against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). base_model = "openai/gpt-5.6-sol" diff --git a/providers/gptzzz/models/gpt-5.6-terra.toml b/providers/gptzzz/models/gpt-5.6-terra.toml index 79ee439c64e..21508e2417a 100644 --- a/providers/gptzzz/models/gpt-5.6-terra.toml +++ b/providers/gptzzz/models/gpt-5.6-terra.toml @@ -1,6 +1,6 @@ -# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live -# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max -# are accepted; minimal is rejected upstream. +# GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the +# live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh +# and max are all accepted on this model; minimal is rejected upstream. # Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked # against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). base_model = "openai/gpt-5.6-terra" diff --git a/providers/gptzzz/models/gpt-5.6.toml b/providers/gptzzz/models/gpt-5.6.toml index 10e2e58d1bf..57eefbd5e85 100644 --- a/providers/gptzzz/models/gpt-5.6.toml +++ b/providers/gptzzz/models/gpt-5.6.toml @@ -1,6 +1,6 @@ -# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live -# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max -# are accepted; minimal is rejected upstream. +# GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the +# live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh +# and max are all accepted on this model; minimal is rejected upstream. # Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked # against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). base_model = "openai/gpt-5.6-sol" diff --git a/providers/gptzzz/models/gpt-6-astra.toml b/providers/gptzzz/models/gpt-6-astra.toml index 17c1dee9685..59764eb4ac6 100644 --- a/providers/gptzzz/models/gpt-6-astra.toml +++ b/providers/gptzzz/models/gpt-6-astra.toml @@ -1,10 +1,11 @@ -# GPTZZZ is an OpenAI-compatible relay. Effort values verified against the live -# /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh, max -# are accepted; minimal is rejected upstream. +# GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the +# live /v1/chat/completions endpoint on 2026-09-16. For gpt-6-astra the host rejects +# none: "Unsupported value: 'none' is not supported with the 'gpt-6-astra' model. +# Supported values are: 'low', 'medium', 'high', 'xhigh', and 'max'." # Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked # against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). base_model = "openai/gpt-6-astra" -reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] +reasoning_options = [{ type = "effort", values = ["low", "medium", "high", "xhigh", "max"] }] [cost] input = 2.00 From 86c56fdad9fac2eee75934e5ab1aa2eb2e493a5e Mon Sep 17 00:00:00 2001 From: pbhhdf <4007625@gmail.com> Date: Mon, 28 Sep 2026 20:37:52 +0800 Subject: [PATCH 3/7] fix(gptzzz): gpt-6-astra now accepts reasoning effort none Re-verified all effort values against the live endpoint on 2026-09-28. gpt-6-astra rejected 'none' on 2026-09-16, which is what this branch recorded. The upstream has since added it: rejecting an unsupported value now reports the supported set as Unsupported value: 'minimal' is not supported with the 'gpt-6-astra' model. Supported values are: 'none', 'low', 'medium', 'high', 'xhigh'. Each value was re-tested per model rather than copied across models. 'minimal' remains rejected everywhere. Co-Authored-By: Claude Opus 5 --- providers/gptzzz/models/gpt-6-astra.toml | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/providers/gptzzz/models/gpt-6-astra.toml b/providers/gptzzz/models/gpt-6-astra.toml index 59764eb4ac6..17e3b301135 100644 --- a/providers/gptzzz/models/gpt-6-astra.toml +++ b/providers/gptzzz/models/gpt-6-astra.toml @@ -1,11 +1,12 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the -# live /v1/chat/completions endpoint on 2026-09-16. For gpt-6-astra the host rejects -# none: "Unsupported value: 'none' is not supported with the 'gpt-6-astra' model. -# Supported values are: 'low', 'medium', 'high', 'xhigh', and 'max'." +# live /v1/chat/completions endpoint, re-verified 2026-09-28. gpt-6-astra rejected none +# on 2026-09-16 but accepts it as of 2026-09-28; the host now reports the supported set +# as "'none', 'low', 'medium', 'high', 'xhigh'" in its rejection of an unsupported value. +# minimal remains rejected on every model. # Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked # against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). base_model = "openai/gpt-6-astra" -reasoning_options = [{ type = "effort", values = ["low", "medium", "high", "xhigh", "max"] }] +reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] [cost] input = 2.00 From e5b06b64981a13a8863ffd0afdfafde0775f9153 Mon Sep 17 00:00:00 2001 From: pbhhdf Date: Tue, 29 Sep 2026 04:17:37 +0800 Subject: [PATCH 4/7] fix(gptzzz): use the public 0.05x group pricing from the model plaza Prices = list price shown on https://gptzzz.ai/model-plaza x 0.05, read from /api/v1/model-plaza on 2026-09-29. List prices for gpt-5.6, gpt-5.6-sol, gpt-5.6-terra and gpt-6-astra were refreshed from the same source. Co-Authored-By: Claude Opus 5.5 --- providers/gptzzz/models/gpt-5.6-sol.toml | 20 ++++++++++---------- providers/gptzzz/models/gpt-5.6-terra.toml | 20 ++++++++++---------- providers/gptzzz/models/gpt-5.6.toml | 20 ++++++++++---------- providers/gptzzz/models/gpt-6-astra.toml | 20 ++++++++++---------- 4 files changed, 40 insertions(+), 40 deletions(-) diff --git a/providers/gptzzz/models/gpt-5.6-sol.toml b/providers/gptzzz/models/gpt-5.6-sol.toml index ae55ff1c1a3..22fd662c4f0 100644 --- a/providers/gptzzz/models/gpt-5.6-sol.toml +++ b/providers/gptzzz/models/gpt-5.6-sol.toml @@ -1,20 +1,20 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the # live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh # and max are all accepted on this model; minimal is rejected upstream. -# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked -# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the +# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). base_model = "openai/gpt-5.6-sol" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] [cost] -input = 0.8 -output = 4.00 -cache_read = 0.08 -cache_write = 1.00 +input = 0.25 +output = 1.5 +cache_read = 0.025 +cache_write = 0.3125 [[cost.tiers]] tier = { size = 272_000 } -input = 1.6 -output = 6.00 -cache_read = 0.16 -cache_write = 2.00 +input = 0.5 +output = 2.25 +cache_read = 0.05 +cache_write = 0.625 diff --git a/providers/gptzzz/models/gpt-5.6-terra.toml b/providers/gptzzz/models/gpt-5.6-terra.toml index 21508e2417a..90b965e6297 100644 --- a/providers/gptzzz/models/gpt-5.6-terra.toml +++ b/providers/gptzzz/models/gpt-5.6-terra.toml @@ -1,20 +1,20 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the # live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh # and max are all accepted on this model; minimal is rejected upstream. -# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked -# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the +# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). base_model = "openai/gpt-5.6-terra" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] [cost] -input = 0.4 -output = 2.4 -cache_read = 0.04 -cache_write = 0.5 +input = 0.1 +output = 0.6 +cache_read = 0.01 +cache_write = 0.125 [[cost.tiers]] tier = { size = 272_000 } -input = 0.8 -output = 3.6 -cache_read = 0.08 -cache_write = 1.00 +input = 0.2 +output = 0.9 +cache_read = 0.02 +cache_write = 0.25 diff --git a/providers/gptzzz/models/gpt-5.6.toml b/providers/gptzzz/models/gpt-5.6.toml index 57eefbd5e85..f64d5a965d6 100644 --- a/providers/gptzzz/models/gpt-5.6.toml +++ b/providers/gptzzz/models/gpt-5.6.toml @@ -1,21 +1,21 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the # live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh # and max are all accepted on this model; minimal is rejected upstream. -# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked -# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the +# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). base_model = "openai/gpt-5.6-sol" name = "GPT-5.6" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] [cost] -input = 0.8 -output = 4.00 -cache_read = 0.08 -cache_write = 1.00 +input = 0.25 +output = 1.5 +cache_read = 0.025 +cache_write = 0.3125 [[cost.tiers]] tier = { size = 272_000 } -input = 1.6 -output = 6.00 -cache_read = 0.16 -cache_write = 2.00 +input = 0.5 +output = 2.25 +cache_read = 0.05 +cache_write = 0.625 diff --git a/providers/gptzzz/models/gpt-6-astra.toml b/providers/gptzzz/models/gpt-6-astra.toml index 17e3b301135..2190b8508ce 100644 --- a/providers/gptzzz/models/gpt-6-astra.toml +++ b/providers/gptzzz/models/gpt-6-astra.toml @@ -3,20 +3,20 @@ # on 2026-09-16 but accepts it as of 2026-09-28; the host now reports the supported set # as "'none', 'low', 'medium', 'high', 'xhigh'" in its rejection of an unsupported value. # minimal remains rejected on every model. -# Prices are the published 0.2x group rate applied to OpenAI list pricing, cross-checked -# against the cost/actual_cost fields returned by https://gptzzz.ai/v1/usage (2026-09-16). +# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the +# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). base_model = "openai/gpt-6-astra" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] [cost] -input = 2.00 -output = 10.00 -cache_read = 0.2 -cache_write = 2.5 +input = 0.5 +output = 2.5 +cache_read = 0.05 +cache_write = 0.625 [[cost.tiers]] tier = { size = 272_000 } -input = 4.00 -output = 15.00 -cache_read = 0.4 -cache_write = 5.00 +input = 1.0 +output = 3.75 +cache_read = 0.1 +cache_write = 1.25 From 61342538b205fca8b2410cba71ce9ec70f02098b Mon Sep 17 00:00:00 2001 From: pbhhdf Date: Tue, 29 Sep 2026 04:18:27 +0800 Subject: [PATCH 5/7] fix(gptzzz): match gpt-6-astra effort values to the host-reported set The host's own rejection message on 2026-09-28 lists 'none', 'low', 'medium', 'high', 'xhigh' for gpt-6-astra, so drop 'max'. Co-Authored-By: Claude Opus 5.5 --- providers/gptzzz/models/gpt-6-astra.toml | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/providers/gptzzz/models/gpt-6-astra.toml b/providers/gptzzz/models/gpt-6-astra.toml index 2190b8508ce..d12a8729e1b 100644 --- a/providers/gptzzz/models/gpt-6-astra.toml +++ b/providers/gptzzz/models/gpt-6-astra.toml @@ -1,12 +1,11 @@ -# GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the -# live /v1/chat/completions endpoint, re-verified 2026-09-28. gpt-6-astra rejected none -# on 2026-09-16 but accepts it as of 2026-09-28; the host now reports the supported set -# as "'none', 'low', 'medium', 'high', 'xhigh'" in its rejection of an unsupported value. -# minimal remains rejected on every model. +# GPTZZZ is an OpenAI-compatible relay. Effort values follow the host's own report for +# gpt-6-astra, re-verified 2026-09-28: probing an unsupported value returns +# "Supported values are: 'none', 'low', 'medium', 'high', 'xhigh'" (no 'max'). +# 'none' was rejected on 2026-09-16 and is accepted as of 2026-09-28; 'minimal' is rejected. # Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the # public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). base_model = "openai/gpt-6-astra" -reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] +reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh"] }] [cost] input = 0.5 From 8c1725db291e59cfc2b0d39cb77cb8b6f9f71333 Mon Sep 17 00:00:00 2001 From: pbhhdf Date: Tue, 29 Sep 2026 04:37:18 +0800 Subject: [PATCH 6/7] fix(gptzzz): state host list prices, drop gpt-6-astra until re-probed gpt-5.6 / gpt-5.6-sol are billed on the host's own list price (5/30 per MTok), which is above OpenAI's; document it in the file headers. Remove gpt-6-astra from this PR; it will come back in a follow-up with per-value probes. Co-Authored-By: Claude Opus 5.5 --- providers/gptzzz/models/gpt-5.6-sol.toml | 7 +++++-- providers/gptzzz/models/gpt-5.6-terra.toml | 5 +++-- providers/gptzzz/models/gpt-5.6.toml | 7 +++++-- providers/gptzzz/models/gpt-6-astra.toml | 21 --------------------- 4 files changed, 13 insertions(+), 27 deletions(-) delete mode 100644 providers/gptzzz/models/gpt-6-astra.toml diff --git a/providers/gptzzz/models/gpt-5.6-sol.toml b/providers/gptzzz/models/gpt-5.6-sol.toml index 22fd662c4f0..446a56a8baa 100644 --- a/providers/gptzzz/models/gpt-5.6-sol.toml +++ b/providers/gptzzz/models/gpt-5.6-sol.toml @@ -1,8 +1,11 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the # live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh # and max are all accepted on this model; minimal is rejected upstream. -# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the -# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). +# Pricing source: https://gptzzz.ai/model-plaza (read from /api/v1/model-plaza on 2026-09-29). +# GPTZZZ bills this model against its own published list price of 5.00 input / 30.00 output / +# 0.50 cache read / 6.25 cache write per MTok (>272K context: 10.00 / 45.00 / 1.00 / 12.50), +# which is higher than OpenAI's list for gpt-5.6-sol, multiplied by its public 0.05x group rate. +# The values below are what the host actually charges. base_model = "openai/gpt-5.6-sol" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] diff --git a/providers/gptzzz/models/gpt-5.6-terra.toml b/providers/gptzzz/models/gpt-5.6-terra.toml index 90b965e6297..73cf2c6a449 100644 --- a/providers/gptzzz/models/gpt-5.6-terra.toml +++ b/providers/gptzzz/models/gpt-5.6-terra.toml @@ -1,8 +1,9 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the # live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh # and max are all accepted on this model; minimal is rejected upstream. -# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the -# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). +# Pricing source: https://gptzzz.ai/model-plaza (read from /api/v1/model-plaza on 2026-09-29). +# Published list price 2.00 / 12.00 / 0.20 / 2.50 per MTok (>272K: 4.00 / 18.00 / 0.40 / 5.00), +# multiplied by the public 0.05x group rate. base_model = "openai/gpt-5.6-terra" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] diff --git a/providers/gptzzz/models/gpt-5.6.toml b/providers/gptzzz/models/gpt-5.6.toml index f64d5a965d6..c58c2c7df99 100644 --- a/providers/gptzzz/models/gpt-5.6.toml +++ b/providers/gptzzz/models/gpt-5.6.toml @@ -1,8 +1,11 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the # live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh # and max are all accepted on this model; minimal is rejected upstream. -# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the -# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). +# Pricing source: https://gptzzz.ai/model-plaza (read from /api/v1/model-plaza on 2026-09-29). +# GPTZZZ bills this model against its own published list price of 5.00 input / 30.00 output / +# 0.50 cache read / 6.25 cache write per MTok (>272K context: 10.00 / 45.00 / 1.00 / 12.50), +# which is higher than OpenAI's list for gpt-5.6-sol, multiplied by its public 0.05x group rate. +# The values below are what the host actually charges. base_model = "openai/gpt-5.6-sol" name = "GPT-5.6" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] diff --git a/providers/gptzzz/models/gpt-6-astra.toml b/providers/gptzzz/models/gpt-6-astra.toml deleted file mode 100644 index d12a8729e1b..00000000000 --- a/providers/gptzzz/models/gpt-6-astra.toml +++ /dev/null @@ -1,21 +0,0 @@ -# GPTZZZ is an OpenAI-compatible relay. Effort values follow the host's own report for -# gpt-6-astra, re-verified 2026-09-28: probing an unsupported value returns -# "Supported values are: 'none', 'low', 'medium', 'high', 'xhigh'" (no 'max'). -# 'none' was rejected on 2026-09-16 and is accepted as of 2026-09-28; 'minimal' is rejected. -# Prices are the list prices published on https://gptzzz.ai/model-plaza multiplied by the -# public 0.05x group rate (read from https://gptzzz.ai/api/v1/model-plaza on 2026-09-29). -base_model = "openai/gpt-6-astra" -reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh"] }] - -[cost] -input = 0.5 -output = 2.5 -cache_read = 0.05 -cache_write = 0.625 - -[[cost.tiers]] -tier = { size = 272_000 } -input = 1.0 -output = 3.75 -cache_read = 0.1 -cache_write = 1.25 From cb3ba8e2873d5f046dd9f47c37360a681db466a1 Mon Sep 17 00:00:00 2001 From: pbhhdf Date: Tue, 29 Sep 2026 04:45:31 +0800 Subject: [PATCH 7/7] fix(gptzzz): re-add gpt-6-astra with probed effort set, add probe and billing evidence Per-value probes on 2026-09-29: gpt-6-astra accepts low..max and rejects none, matching the OpenAI lab entry. gpt-5.6* accept none..max. /v1/usage cost matches the stated list prices and actual_cost/cost = 0.05. Co-Authored-By: Claude Opus 5.5 --- providers/gptzzz/models/gpt-5.6-sol.toml | 7 ++++--- providers/gptzzz/models/gpt-5.6-terra.toml | 6 +++--- providers/gptzzz/models/gpt-5.6.toml | 7 ++++--- providers/gptzzz/models/gpt-6-astra.toml | 24 ++++++++++++++++++++++ 4 files changed, 35 insertions(+), 9 deletions(-) create mode 100644 providers/gptzzz/models/gpt-6-astra.toml diff --git a/providers/gptzzz/models/gpt-5.6-sol.toml b/providers/gptzzz/models/gpt-5.6-sol.toml index 446a56a8baa..a803e00af59 100644 --- a/providers/gptzzz/models/gpt-5.6-sol.toml +++ b/providers/gptzzz/models/gpt-5.6-sol.toml @@ -1,11 +1,12 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the -# live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh -# and max are all accepted on this model; minimal is rejected upstream. +# live /v1/chat/completions endpoint on 2026-09-16 and re-probed on 2026-09-29: none, low, +# medium, high, xhigh and max all return HTTP 200; minimal returns HTTP 400. # Pricing source: https://gptzzz.ai/model-plaza (read from /api/v1/model-plaza on 2026-09-29). # GPTZZZ bills this model against its own published list price of 5.00 input / 30.00 output / # 0.50 cache read / 6.25 cache write per MTok (>272K context: 10.00 / 45.00 / 1.00 / 12.50), # which is higher than OpenAI's list for gpt-5.6-sol, multiplied by its public 0.05x group rate. -# The values below are what the host actually charges. +# The host's /v1/usage "cost" field matches this 5.00 / 30.00 list exactly, so the values +# below are what the host actually charges. base_model = "openai/gpt-5.6-sol" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] diff --git a/providers/gptzzz/models/gpt-5.6-terra.toml b/providers/gptzzz/models/gpt-5.6-terra.toml index 73cf2c6a449..65e402b588a 100644 --- a/providers/gptzzz/models/gpt-5.6-terra.toml +++ b/providers/gptzzz/models/gpt-5.6-terra.toml @@ -1,9 +1,9 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the -# live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh -# and max are all accepted on this model; minimal is rejected upstream. +# live /v1/chat/completions endpoint on 2026-09-16 and re-probed on 2026-09-29: none, low, +# medium, high, xhigh and max all return HTTP 200; minimal returns HTTP 400. # Pricing source: https://gptzzz.ai/model-plaza (read from /api/v1/model-plaza on 2026-09-29). # Published list price 2.00 / 12.00 / 0.20 / 2.50 per MTok (>272K: 4.00 / 18.00 / 0.40 / 5.00), -# multiplied by the public 0.05x group rate. +# multiplied by the public 0.05x group rate. The host's /v1/usage "cost" field matches this list. base_model = "openai/gpt-5.6-terra" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] diff --git a/providers/gptzzz/models/gpt-5.6.toml b/providers/gptzzz/models/gpt-5.6.toml index c58c2c7df99..02ec739c8e5 100644 --- a/providers/gptzzz/models/gpt-5.6.toml +++ b/providers/gptzzz/models/gpt-5.6.toml @@ -1,11 +1,12 @@ # GPTZZZ is an OpenAI-compatible relay. Effort values verified per model against the -# live /v1/chat/completions endpoint on 2026-09-16: none, low, medium, high, xhigh -# and max are all accepted on this model; minimal is rejected upstream. +# live /v1/chat/completions endpoint on 2026-09-16 and re-probed on 2026-09-29: none, low, +# medium, high, xhigh and max all return HTTP 200; minimal returns HTTP 400. # Pricing source: https://gptzzz.ai/model-plaza (read from /api/v1/model-plaza on 2026-09-29). # GPTZZZ bills this model against its own published list price of 5.00 input / 30.00 output / # 0.50 cache read / 6.25 cache write per MTok (>272K context: 10.00 / 45.00 / 1.00 / 12.50), # which is higher than OpenAI's list for gpt-5.6-sol, multiplied by its public 0.05x group rate. -# The values below are what the host actually charges. +# The host's /v1/usage "cost" field matches this 5.00 / 30.00 list exactly, so the values +# below are what the host actually charges. base_model = "openai/gpt-5.6-sol" name = "GPT-5.6" reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }] diff --git a/providers/gptzzz/models/gpt-6-astra.toml b/providers/gptzzz/models/gpt-6-astra.toml new file mode 100644 index 00000000000..cbc0c8dee38 --- /dev/null +++ b/providers/gptzzz/models/gpt-6-astra.toml @@ -0,0 +1,24 @@ +# GPTZZZ is an OpenAI-compatible relay. Effort values probed one by one against the live +# /v1/chat/completions endpoint on 2026-09-29: low, medium, high, xhigh and max return +# HTTP 200 (reasoning_tokens 10 / 11 / 14 at high / xhigh / max); none and minimal return +# HTTP 400 ("Supported values are: 'low', 'medium', 'high', 'xhigh', and 'max'"). +# This matches the OpenAI lab entry. +# Pricing source: https://gptzzz.ai/model-plaza (read from /api/v1/model-plaza on 2026-09-29). +# Published list price 10.00 / 50.00 / 1.00 / 12.50 per MTok (>272K: 20.00 / 75.00 / 2.00 / 25.00), +# multiplied by the public 0.05x group rate. The host's /v1/usage "cost" field matches this +# list price exactly. +base_model = "openai/gpt-6-astra" +reasoning_options = [{ type = "effort", values = ["low", "medium", "high", "xhigh", "max"] }] + +[cost] +input = 0.5 +output = 2.5 +cache_read = 0.05 +cache_write = 0.625 + +[[cost.tiers]] +tier = { size = 272_000 } +input = 1.0 +output = 3.75 +cache_read = 0.1 +cache_write = 1.25